{"id":"2b3c6722-25ec-447f-a499-a9f8088b9efd","entityType":"agent","slug":"clawhub-cellcog-cellcog","name":"cellcog","canonicalUrl":"https://www.xpersona.co/agent/clawhub-cellcog-cellcog","canonicalPath":"/agent/clawhub-cellcog-cellcog","generatedAt":"2026-10-09T10:36:07.317Z","source":"CLAWHUB","claimStatus":"UNCLAIMED","verificationTier":"NONE","summary":{"evidence":{"source":"editorial-content","verified":true,"confidence":"high","updatedAt":"2026-10-09T01:46:45.227Z","emptyReason":null},"description":"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code. Skill: cellcog Owner: cellcog Summary: Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code. Tags: latest","descriptionLabel":"Technical summary","evidenceSummary":"Capability contract not published. No trust telemetry is available yet. 18.4K downloads reported by the source. Last updated 10/9/2026.","installCommand":"clawhub skill install s176q1btpn094ats9b4f9hgfwx83knpk:cellcog","sourceUrl":"https://clawhub.ai/cellcog/cellcog","homepage":"https://clawhub.ai/cellcog/skills/cellcog","primaryLinks":[{"label":"View on ClawHub","url":"https://clawhub.ai/cellcog/cellcog","kind":"source"},{"label":"Homepage","url":"https://clawhub.ai/cellcog/skills/cellcog","kind":"homepage"}],"safetyScore":84,"overallRank":62,"popularityScore":85,"trustScore":null,"claimedByName":null,"isOwner":false,"seoDescription":"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and"},"coverage":{"evidence":{"source":"public-profile","verified":false,"confidence":"medium","updatedAt":"2026-10-09T01:46:45.227Z","emptyReason":null},"protocols":[{"protocol":"OPENCLEW","label":"OpenClaw","status":"self-declared","notes":"Declared in the public agent profile."}],"capabilities":[],"verifiedCount":0,"selfDeclaredCount":1,"capabilityMatrix":{"rows":[{"key":"OPENCLEW","type":"protocol","support":"unknown","confidenceSource":"profile","notes":"Listed on profile"}],"flattenedTokens":"protocol:OPENCLEW|unknown|profile"}},"adoption":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-09T01:46:45.227Z","emptyReason":null},"stars":null,"forks":null,"downloads":18393,"packageName":null,"latestVersion":"2.0.21","tractionLabel":"18.4K downloads"},"release":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-09T01:46:45.227Z","emptyReason":null},"lastUpdatedAt":"2026-10-09T01:46:45.227Z","lastCrawledAt":"2026-10-09T01:46:45.227Z","lastIndexedAt":null,"nextCrawlAt":"2026-10-10T01:46:45.227Z","lastVerifiedAt":null,"highlights":[{"version":"2.0.21","createdAt":"2026-08-24T01:51:01.035Z","changelog":"Display title updated.","fileCount":3,"zipByteSize":8905},{"version":"2.0.20","createdAt":"2026-08-24T01:45:14.207Z","changelog":"Title and content updated.","fileCount":3,"zipByteSize":8905},{"version":"2.0.19","createdAt":"2026-08-09T15:19:31.179Z","changelog":"Display title updated.","fileCount":3,"zipByteSize":7952},{"version":"2.0.18","createdAt":"2026-08-03T06:05:40.283Z","changelog":"Content updated.","fileCount":3,"zipByteSize":7899},{"version":"2.0.17","createdAt":"2026-07-22T15:57:11.610Z","changelog":"- Documentation updated in SKILL.md; content revised or expanded. - Obsolete or redundant file skill-card.md removed.","fileCount":3,"zipByteSize":8010},{"version":"2.0.16","createdAt":"2026-07-17T03:40:30.425Z","changelog":"- Updated DeepResearch Bench ranking reference to July 2026 and clarified that leaderboard positions may change over time. - Removed stray future-dated leaderboard claim (\"#1 on DeepResearch Bench (Apr 2026)\") in favor of more accurate, dynamic wording. - Removed the file skill-card.md. - No behavioral changes; documentation refresh only.","fileCount":3,"zipByteSize":7862},{"version":"2.0.15","createdAt":"2026-04-23T07:02:12.615Z","changelog":"- Documentation clarifies that anything wrapped in tags is uploaded to CellCog; warns not to include credentials, private keys, or sensitive material as SHOW_FILE files. - Added a comparison to reference image attachments in Nano Banana to explain usage. - Installation section now specifies that cellcog is the official Python SDK, with source and PyPI links. - Wait/Notify mode comparison table updated: Notify mode is now OpenClaw-only, not available for other agents; clarified table language and guidance. - Minor edits for accuracy and improved guidance throughout the setup and usage sections; small formatting improvements.","fileCount":3,"zipByteSize":7685},{"version":"2.0.14","createdAt":"2026-04-15T02:33:14.825Z","changelog":"- Added support for new modalities: avatars and voice cloning in the SKILL description. - Updated description to reflect broader capabilities (now includes avatars and voice cloning in addition to existing formats). - No code or API changes—documentation/feature list update only.","fileCount":2,"zipByteSize":6282}]},"execution":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No published capability contract is available yet."},"installCommand":"clawhub skill install s176q1btpn094ats9b4f9hgfwx83knpk:cellcog","setupComplexity":"low","setupSteps":["Setup complexity is classified as HIGH. You must provision dedicated cloud infrastructure or an isolated VM. Do not run this directly on your local workstation.","Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data."],"contract":{"contractStatus":"missing","authModes":[],"requires":[],"forbidden":[],"supportsMcp":false,"supportsA2a":false,"supportsStreaming":false,"inputSchemaRef":null,"outputSchemaRef":null,"dataRegion":null,"contractUpdatedAt":null,"sourceUpdatedAt":null,"freshnessSeconds":null},"invocationGuide":{"preferredApi":{"snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-cellcog-cellcog/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-cellcog-cellcog/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-cellcog-cellcog/trust"},"curlExamples":["curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-cellcog-cellcog/snapshot\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-cellcog-cellcog/contract\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-cellcog-cellcog/trust\""],"jsonRequestTemplate":{"query":"summarize this repo","constraints":{"maxLatencyMs":2000,"protocolPreference":["OPENCLEW"]}},"jsonResponseTemplate":{"ok":true,"result":{"summary":"...","confidence":0.9},"meta":{"source":"CLAWHUB","generatedAt":"2026-10-09T10:36:07.313Z"}},"retryPolicy":{"maxAttempts":3,"backoffMs":[500,1500,3500],"retryableConditions":["HTTP_429","HTTP_503","NETWORK_TIMEOUT"]}},"endpoints":{"dossierUrl":"https://www.xpersona.co/api/v1/agents/clawhub-cellcog-cellcog/dossier","snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-cellcog-cellcog/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-cellcog-cellcog/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-cellcog-cellcog/trust"}},"reliability":{"evidence":{"source":"runtime-metrics","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No trust, reliability, or runtime telemetry is available."},"trust":{"status":"unavailable","handshakeStatus":"UNKNOWN","verificationFreshnessHours":null,"reputationScore":null,"p95LatencyMs":null,"successRate30d":null,"fallbackRate":null,"attempts30d":null,"trustUpdatedAt":null,"trustConfidence":"unknown","sourceUpdatedAt":null,"freshnessSeconds":null},"decisionGuardrails":{"doNotUseIf":["Contract metadata is missing or unavailable for deterministic execution."],"safeUseWhen":[],"riskFlags":["missing_or_unavailable_contract","trust_data_unavailable","schema_references_missing"],"operationalConfidence":"low"},"executionMetrics":{"observedLatencyMsP50":null,"observedLatencyMsP95":null,"estimatedCostUsd":null,"uptime30d":null,"rateLimitRpm":null,"rateLimitBurst":null,"lastVerifiedAt":null,"verificationSource":null},"runtimeMetrics":{"successRate":null,"avgLatencyMs":null,"avgCostUsd":null,"hallucinationRate":null,"retryRate":null,"disputeRate":null,"p50Latency":null,"p95Latency":null,"lastUpdated":null}},"benchmarks":{"evidence":{"source":"no-benchmark-data","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No benchmark suites or observed failure patterns are available."},"suites":[],"failurePatterns":[]},"artifacts":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"high","updatedAt":"2026-10-09T01:46:45.227Z","emptyReason":null},"readme":"Skill: cellcog\n\nOwner: cellcog\n\nSummary: Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code.\n\nTags: latest:2.0.21\n\nVersion history:\n\nv2.0.21 | 2026-08-24T01:51:01.035Z | user\n\nDisplay title updated.\n\nv2.0.20 | 2026-08-24T01:45:14.207Z | user\n\nTitle and content updated.\n\nv2.0.19 | 2026-08-09T15:19:31.179Z | user\n\nDisplay title updated.\n\nv2.0.18 | 2026-08-03T06:05:40.283Z | user\n\nContent updated.\n\nv2.0.17 | 2026-07-22T15:57:11.610Z | auto\n\n- Documentation updated in SKILL.md; content revised or expanded.\n- Obsolete or redundant file skill-card.md removed.\n\nv2.0.16 | 2026-07-17T03:40:30.425Z | auto\n\n- Updated DeepResearch Bench ranking reference to July 2026 and clarified that leaderboard positions may change over time.\n- Removed stray future-dated leaderboard claim (\"#1 on DeepResearch Bench (Apr 2026)\") in favor of more accurate, dynamic wording.\n- Removed the file skill-card.md.\n- No behavioral changes; documentation refresh only.\n\nv2.0.15 | 2026-04-23T07:02:12.615Z | auto\n\n- Documentation clarifies that anything wrapped in tags is uploaded to CellCog; warns not to include credentials, private keys, or sensitive material as SHOW_FILE files.\n- Added a comparison to reference image attachments in Nano Banana to explain usage.\n- Installation section now specifies that cellcog is the official Python SDK, with source and PyPI links.\n- Wait/Notify mode comparison table updated: Notify mode is now OpenClaw-only, not available for other agents; clarified table language and guidance.\n- Minor edits for accuracy and improved guidance throughout the setup and usage sections; small formatting improvements.\n\nv2.0.14 | 2026-04-15T02:33:14.825Z | auto\n\n- Added support for new modalities: avatars and voice cloning in the SKILL description.\n- Updated description to reflect broader capabilities (now includes avatars and voice cloning in addition to existing formats).\n- No code or API changes—documentation/feature list update only.\n\nv2.0.13 | 2026-04-14T17:31:42.197Z | auto\n\n- Expanded description to emphasize agent-to-agent protocol and multi-step iteration for high accuracy.\n- Updated and clarified introductory sections to stress support for all output types, not just code.\n- Improved highlights of CellCog's DeepResearch Bench #1 ranking and modal capabilities.\n- Cleaned up and streamlined the documentation for easier understanding without removing key usage instructions.\n- Clarified that CellCog ensures all your work gets done, in any modality.\n\nv2.0.12 | 2026-04-13T01:00:20.478Z | auto\n\n- Updated the installation instruction to use `pip install -U cellcog` (was `pip install cellcog`).\n- No functional or API changes introduced in this update.\n- Documentation remains focused on setup, usage, and agent integration.\n\nv2.0.11 | 2026-04-12T23:34:54.798Z | auto\n\n- SKILL.md significantly shortened and streamlined.\n- Description made more concise while maintaining core features and DeepResearch Bench claim.\n- Long, detailed explanations and examples were retained, but general wording and metadata were made briefer.\n- Content after the \"Response Shape\" heading was cut off for brevity.\n- No changes to skill logic or API; documentation update only.\n\nv2.0.10 | 2026-04-11T05:52:25.505Z | auto\n\n- Added requirement to specify agent_provider when initializing CellCogClient (e.g., \"openclaw\", \"cursor\", \"claude-code\", \"aider\", \"windsurf\", \"perplexity\", \"hermes\", \"script\").\n- Updated documentation examples, replacing agent_name with agent_provider for clarity and correct usage.\n- Improved onboarding instructions for better compatibility across agent frameworks.\n\nv2.0.9 | 2026-04-11T02:58:37.618Z | auto\n\n- Clarified SDK usage by specifying agent_name=\"openclaw\" in code examples for both OpenClaw and non-OpenClaw agents.\n- Updated description to remove \"deep research\" to improve clarity.\n- No other functionality or interface changes.\n\nv2.0.8 | 2026-04-08T05:52:22.595Z | auto\n\n- Updated description to highlight deep research as a primary use case.\n- Minor text edits for clarity and emphasis in SKILL.md.\n- No functional or code-level changes; documentation only.\n\nv2.0.7 | 2026-04-06T05:06:45.388Z | auto\n\ncellcog 2.0.7 brings a streamlined description and onboarding for the #1 Any-to-Any agent sub-agent.\n\n- Updated SKILL.md with a concise, clearer description highlighting Any-to-Any capability, modal coverage, and high-accuracy agent-to-agent protocol.\n- Refined install/setup guidance and emphasized practical SDK use for research, documents, images, audio, video, code, diagrams, and more.\n- Improved positioning for sub-agent and multi-step workflows, clarifying CellCog's value beyond code output.\n- No code or API changes in this version; documentation only.\n\nv2.0.6 | 2026-04-03T01:42:58.046Z | auto\n\n- Clarified how to use tags for file inputs; added explicit DOs and DON'Ts.\n- Added guidance on explicitly requesting output artifacts to ensure correct deliverables.\n- Separated OpenClaw (notify) and universal (wait) usage sections with clearer instructions and mode selection table.\n- Enhanced explanations on when to use \"notify\" vs \"wait\" workflow for various agent scenarios.\n- Improved warnings on always printing the full response message.\n- General documentation restructuring and clarification for better usability.\n\nv2.0.5 | 2026-04-03T00:07:21.646Z | auto\n\nCellCog 2.0.5 – Revised API Usage & Docs\n\n- Updated SKILL.md with new task creation patterns and clearer guidance for both \"wait for completion\" (blocking, universal) and \"notify on completion\" (OpenClaw/daemon) modes.\n- Clarified SDK response shape; emphasized always printing the full message for complete outputs and follow-ups.\n- Improved timeout and resumption handling instructions.\n- Added more details for optional parameters and the notification system.\n- Removed SKILL_original.md (now redundant).\n\nv2.0.4 | 2026-04-02T05:54:41.089Z | auto\n\n- Documentation updated in SKILL.md.\n- No functional or API changes; content only.\n- Existing examples and instructions retained.\n- No impact to skill operation or user integrations.\n\nv2.0.3 | 2026-04-02T04:25:56.354Z | auto\n\n- Documentation formatting updated in SKILL.md to improve readability.\n- No functional or code changes; content remains the same.\n- No new features or bug fixes included.\n\nv2.0.2 | 2026-04-02T03:33:28.461Z | auto\n\n- Updated DeepResearch Bench reference from February 2026 to April 2026 for all leaderboard mentions.\n- Clarified and simplified instructions for file path usage and output requests.\n- Streamlined credit usage explanation, noting orchestration of 21+ foundation models and removal of prior credit estimate discussion.\n- Expanded documentation for task creation, including all optional parameters and required fields.\n- Added a detailed \"Chat Modes\" section describing available modes, their use cases, speed, and credit requirements.\n- Improved guidance and examples for session key usage and workflow instructions.\n\nv2.0.1 | 2026-03-31T20:12:14.497Z | auto\n\n- Documentation in SKILL.md has been updated.\n- No changes to core functionality or interface.\n\nv2.0.0 | 2026-03-27T04:07:13.126Z | auto\n\ncellcog 2.0.0\n\n- Added `homepage` field with the project website.\n- Refactored and expanded metadata: clarified supported operating systems and environment variables, and added structured install instructions.\n- Improved standardization and clarity in the SKILL.md header for better integration and discoverability.\n- No changes to core usage documentation or API content.\n\nv1.0.26 | 2026-03-26T06:47:17.389Z | auto\n\ncellcog 1.0.26\n\n- Expanded API reference for create_chat(): now documents optional project_id and agent_role_id parameters for advanced workflows.\n- No changes to code or behavior; update improves user guidance and advanced usage documentation.\n\nv1.0.25 | 2026-03-25T23:56:23.283Z | auto\n\n- Removed the \"Typical Credit Costs\" table; documentation now explicitly states that credit usage estimates are not provided, with an explanation of why.\n- Updated the \"Quick Start\" section to clarify how credit consumption works and to set expectations for users regarding variability.\n- No changes to API or functionality; documentation update only.\n\nv1.0.24 | 2026-03-23T21:52:47.044Z | auto\n\n- Documentation updated in SKILL.md; no functional code changes.\n- Ensures that instructions, usage examples, and explanations remain consistent and clear for all users.\n\nv1.0.23 | 2026-03-19T11:51:36.295Z | user\n\ncellcog 1.0.23\n\n- Added documentation for wait_for_completion(), enabling agents to block until a CellCog task finishes.\n- Clarified guidance on using wait_for_completion() in automated workflows like cron jobs or pipelines.\n- Minor text and formatting improvements in usage instructions.\n- No code changes; update affects documentation only.\n\nv1.0.22 | 2026-03-19T02:51:29.436Z | user\n\ncellcog 1.0.22\n\n- Updated documentation to add guidance and cost estimates for the new \"Agent Team Max\" mode.\n- Expanded typical credit cost table to include \"Deep research (Agent Team Max mode)\".\n- Clarified cost multipliers for Agent Team and Agent Team Max modes.\n- Updated Chat Modes table and descriptions to reflect support and requirements for \"agent team max\".\n\nv1.0.21 | 2026-02-21T18:00:56.727Z | auto\n\n- Documentation (SKILL.md) updated for clarity and completeness.\n- No changes to code or functionality.\n\nv1.0.20 | 2026-02-17T23:00:09.697Z | user\n\nNo code or documentation changes detected in this version.\n\nv1.0.19 | 2026-02-17T20:45:40.756Z | user\n\nNo changes detected for version 1.0.19.\n\n- No file changes were found in this release.\n- Functionality and documentation remain the same as the previous version.\n\nv1.0.18 | 2026-02-17T20:19:31.737Z | user\n\nVersion 1.0.18\n\n- Removed detailed \"Account Setup\" and credit purchase instructions from documentation.\n- Now references typical credit costs directly, making it easier to estimate usage requirements.\n- Streamlined setup and onboarding details for a quicker start.\n- No changes detected in code or file structure; documentation update only.\n\nv1.0.17 | 2026-02-17T02:42:13.922Z | user\n\n- Added detailed information about CellCog's required credit system—API key and credits are now both necessary to use the service.\n- Included credit cost estimates for common task types and provided plan recommendations tailored to varying usage levels.\n- Updated account setup instructions, directing users to set up both API keys and credits via the CellCog website.\n- Clarified notification structure and emphasized the importance of the \"Why\" section in task completion messages.\n- No code or API changes; documentation and onboarding instructions improved for clarity regarding billing and usage.\n\nv1.0.16 | 2026-02-14T10:15:10.303Z | user\n\nNo user-visible changes in this version. \n- No file changes detected.\n- Documentation and functionality remain unchanged from previous release.\n\nv1.0.15 | 2026-02-11T01:35:39.109Z | user\n\n**New environment variable and chat deletion added for improved setup and privacy.**\n\n- Added support for configuring API keys via the `CELLCOG_API_KEY` environment variable (recommended setup).\n- Introduced `delete_chat()` API to permanently delete chats and their data from CellCog servers.\n- Updated documentation to clarify file path requirements (absolute, within `` tags).\n- Minor clarification and formatting improvements in usage examples and authentication steps.\n\nv1.0.14 | 2026-02-11T01:09:28.059Z | user\n\n- Removed author and environment variable information from metadata.\n- Updated authentication section: now uses explicit client.set_api_key() instead of relying on environment variables.\n- Cleaned up and streamlined installation and setup instructions.\n- Removed delete_chat() API method from documentation.\n- Minor clarifications and edits for conciseness and clarity throughout the skill documentation.\n\nv1.0.13 | 2026-02-11T01:05:23.465Z | auto\n\n- Added author and platform metadata fields (`author`, `env`, `os`, `install`) for improved skill packaging and clarity.\n- Recommended authentication via the `CELLCOG_API_KEY` environment variable instead of only in-code setting.\n- Minor documentation corrections and clarifications (e.g., fixed \"abs\" → \"absolute\" regarding file paths, and improved clarity about import/install instructions).\n- Documented a new API method: `delete_chat()`, including guidance on secure, server-side deletion of chat data.\n- Overall structure and best practices updated for easier installation and improved security guidance.\n\nv1.0.12 | 2026-02-09T20:48:08.163Z | user\n\n- No code or documentation changes detected in this version.\n- Version number updated only; functionality and documentation remain unchanged.\n\nv1.0.11 | 2026-02-09T20:01:16.261Z | user\n\nNo user-visible changes in this release.\n\n- Version updated to 1.0.11 with no modifications to files or documentation.\n\nv1.0.10 | 2026-02-09T04:19:10.776Z | user\n\n- No file changes detected in this release.\n- Documentation and usage information remain unchanged from the previous version.\n- No new features, fixes, or modifications introduced in this update.\n\nv1.0.9 | 2026-02-08T19:05:21.942Z | auto\n\ncellcog 1.0.9\n\n- Documentation updated to clarify that file paths must be absolute and enclosed in tags when referencing files in CellCog prompts.\n- Minor formatting changes for better readability in the usage examples.\n\nv1.0.8 | 2026-02-08T17:52:19.095Z | user\n\n- No changes detected in this version.\n- All features, documentation, and instructions remain the same.\n\nv1.0.7 | 2026-02-08T00:42:11.504Z | auto\n\n- Documentation updates in SKILL.md for improved clarity.\n- Section \"Send Multiple Files, Any Format\" retitled to \"Work With Multiple Files, Any Format\".\n- Minor phrasing adjustments for conciseness and readability.\n- No breaking changes to features or API.\n\nv1.0.6 | 2026-02-07T02:12:20.567Z | user\n\n- No user-facing or backend file changes detected in this release.\n- All features, API, and documentation remain unchanged from the previous version.\n\nv1.0.5 | 2026-02-06T20:51:00.730Z | auto\n\n- Skill description is now more concise and highlights DeepResearch Bench ranking in the first line.\n- Minor language improvements and clarifications in the intro sections.\n- No code/API or functional changes—documentation only.\n- No breaking changes; usage and API remain identical.\n\nv1.0.4 | 2026-02-06T20:39:46.131Z | auto\n\n**Summary:** Refined documentation for clarity and added recent performance claims.\n\n- Improved and condensed setup instructions; simplified SDK installation and authentication steps.\n- Emphasized CellCog’s deep reasoning capabilities and included performance ranking (#1 on DeepResearch Bench as of Feb 2026).\n- Clarified real-time progress update and notification behaviors for long-running tasks.\n- Streamlined API usage instructions and core method references.\n- Added clearer explanation of chat modes and reasoning benefits.\n- Removed older implementation details and reduced repetition.\n\nv1.0.3 | 2026-02-06T20:03:30.248Z | user\n\ncellcog v1.0.3\n\n- Updated documentation to clarify that agent mode typically asks clarifying questions within 1-2 minutes.\n- Added instructions for skipping clarifying questions by including \"No clarifying questions needed\" in the prompt.\n- Improved explanation messages in sample outputs and return values to highlight clarifying question behavior.\n- General documentation refinements for clarity and accuracy regarding notification and response timing.\n\nv1.0.2 | 2026-02-05T05:48:22.625Z | user\n\n**CellCog v1.0.2 — Switch to Fire-and-Forget Execution Pattern**\n\n- Adopts a \"fire-and-forget\" workflow using background WebSocket notifications for task completion—no need to spawn or manage sub-sessions.\n- Updated usage instructions to showcase immediate returns from `create_chat()`/`send_message()`, and daemon-based result delivery to your session.\n- Adds detailed explanations on task notifications, interim progress updates, and full-result delivery to streamline integration.\n- Now explicitly notes SDK version compatibility and offers troubleshooting instructions for SDK mismatch.\n- Minor enhancements to documentation clarity, table formatting, and skill metadata (including emoji).\n\nv1.0.1 | 2026-02-04T03:37:03.445Z | user\n\n**Summary:** Major documentation overhaul for clarity, focus, and next-generation Any-to-Any workflow.\n\n- Completely rewrote documentation for clarity and conciseness.\n- Focused on CellCog’s unique Any-to-Any capabilities with practical, multi-modal examples.\n- Streamlined quick start, installation, authentication, and session management steps.\n- Improved code samples showing session spawning, streaming, and file handling patterns.\n- Added straightforward error and advanced task management sections.\n- Renamed and clarified references to sub-skills (e.g., research-cog, video-cog, etc.).\n- Removed overly verbose explanations; condensed core usage and integration guidance.\n\nv1.0.0 | 2026-02-03T07:36:30.626Z | user\n\nCellCog 1.0.0 - Initial Release\n\n- Introduces CellCog: a sub-agent for high-quality, complex, and multimodal tasks across conversational AI, research, and a wide range of input/output types.\n- Provides a unified API to handle text, images, videos, audio, documents, code, and more, supporting advanced deliverables (reports, dashboards, presentations, etc.).\n- Enforces always using spawned sessions (`sessions_spawn`) to prevent blocking the main agent and enable parallel task processing.\n- Includes comprehensive documentation: setup instructions, usage patterns, best practices, and complete code examples.\n- Requires a CellCog API key for activation; simple SDK integration with automated file handling.\n\nArchive index:\n\nArchive v2.0.21: 3 files, 8905 bytes\n\nFiles: skill-card.md (2439b), SKILL.md (18222b), _meta.json (127b)\n\nFile v2.0.21:SKILL.md\n\n---\nname: cellcog\ndescription: \"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code.\"\nauthor: CellCog\nhomepage: https://cellcog.ai\nmetadata:\n openclaw:\n emoji: \"🧠\"\n os: [darwin, linux, windows]\n requires:\n bins: [python3]\n env: [CELLCOG_API_KEY]\n\n---\n# CellCog - Any-to-Any for Agents\n\n## The Power of Any-to-Any\n\nCellCog is the only AI that truly handles **any input → any output** in a single request. No tool chaining. No orchestration complexity. One call, multiple deliverables.\n\nCellCog pairs all modalities with frontier-level deep reasoning — as of July 2026, CellCog is **#1 on the DeepResearch Bench** (rankings change frequently; see the live leaderboard for the latest): https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard\n\n### Work With Multiple Files, Any Format\n\nReference as many documents as you need—all at once:\n\n```python\nprompt = \"\"\"\nAnalyze all of these together:\n/data/q4_earnings.pdf\n/data/competitor_analysis.pdf\n/data/market_research.xlsx\n/recordings/customer_interview.mp3\n/designs/product_mockup.png\n\nGive me a comprehensive market positioning analysis based on all these inputs.\n\"\"\"\n```\n\nFile paths must be absolute and enclosed in `` tags. CellCog understands PDFs, spreadsheets, images, audio, video, code files, and more.\n\n⚠️ **Without SHOW_FILE tags, CellCog only sees the path as text — not the file contents.**\n\n❌ `Analyze /data/sales.csv` — CellCog can't read the file\n✅ `Analyze /data/sales.csv` — CellCog reads it\n\n### Think of SHOW_FILE like reference files\n\nJust like Nano Banana accepts reference images, CellCog accepts reference files of any type — PDFs, spreadsheets, audio, code, images — as inputs the model reads during a task. Same mental model as any multimodal AI attachment.\n\n**Only attach what you intend to share.** Anything inside a `` tag is uploaded to CellCog. Don't wrap credentials, private keys, `.env` files, SSH keys, or other sensitive material in SHOW_FILE tags — the same way you wouldn't paste them into Nano Banana, ChatGPT, or any other AI service's file upload.\n\n### Request Multiple Outputs, Different Modalities\n\nAsk for completely different output types in ONE request:\n\n```python\nprompt = \"\"\"\nBased on this quarterly sales data:\n/data/sales_q4_2025.csv\n\nCreate ALL of the following:\n1. A PDF executive summary report with charts\n2. An interactive HTML dashboard for the leadership team\n3. A 60-second video presentation for the all-hands meeting\n4. A slide deck for the board presentation\n5. An Excel file with the underlying analysis and projections\n\"\"\"\n```\n\nCellCog handles the entire workflow — analyzing, generating, and delivering all outputs with consistent insights across every format.\n\n⚠️ **Be explicit about output artifacts.** Without explicit artifact language, CellCog may respond with text analysis instead of generating a file.\n\n❌ `\"Quarterly earnings analysis for AAPL\"` — could produce text or any format\n✅ `\"Create a PDF report and an interactive HTML dashboard analyzing AAPL quarterly earnings.\"` — CellCog creates actual deliverables\n\n**Your sub-agent for quality work.** Depth, accuracy, and real deliverables.\n\n---\n\n## Quick Start\n\n### Setup\n\n```python\nfrom cellcog import CellCogClient\n```\n\nIf import fails, install the official CellCog Python SDK:\n```bash\npip install -U cellcog\n```\n\n`cellcog` is the official Python SDK maintained by CellCog AI Inc. Source: https://github.com/CellCog/cellcog_python · Package: https://pypi.org/project/cellcog/\n\n### Authentication\n\n**Environment variable (recommended):** Set `CELLCOG_API_KEY` — the SDK picks it up automatically:\n```bash\nexport CELLCOG_API_KEY=\"sk_...\"\n```\n\nGet API key from: https://cellcog.ai/profile?tab=api-keys\n\n```python\nstatus = client.get_account_status()\nprint(status) # {\"configured\": True, \"email\": \"user@example.com\", ...}\n```\n\n### Agent Provider\n\n`agent_provider` is **required** when creating a `CellCogClient`. It identifies which agent framework is calling CellCog — not your individual agent's name, but the platform/tool you're running inside.\n\nExamples: `\"openclaw\"`, `\"claude-code\"`, `\"cursor\"`, `\"aider\"`, `\"windsurf\"`, `\"perplexity\"`, `\"hermes\"`, `\"script\"` (for standalone scripts).\n\n### OpenClaw Agents\n\nFire-and-forget — your agent stays free while CellCog works:\n\n```python\nclient = CellCogClient(agent_provider=\"openclaw\")\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n notify_session_key=\"agent:main:main\", # OpenClaw session key\n task_label=\"quantum-research\", # Label for notifications\n chat_mode=\"agent\",\n)\n# Returns IMMEDIATELY — daemon delivers results to your session when done\n```\n\n### All Other Agents (Cursor, Claude Code, etc.)\n\nBlocks until done — simplest pattern:\n\n```python\nclient = CellCogClient(agent_provider=\"cursor\") # or \"claude-code\", \"aider\", \"script\", etc.\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n task_label=\"quantum-research\",\n chat_mode=\"agent\",\n)\n# Blocks until done — result contains everything\nprint(result[\"message\"])\n```\n\n### Credit Usage\n\nCellCog orchestrates 21+ frontier foundation models. Credit consumption is unpredictable and varies by task complexity. Credits used are reported in every completion notification.\n\n---\n\n## Creating Tasks\n\n### Notify on Completion (OpenClaw — Fire-and-Forget)\n\nReturns immediately. A background daemon monitors via WebSocket and delivers results to your session when done. Your agent stays free to take new instructions, start other tasks, or continue working.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n notify_session_key=\"agent:main:main\", # Required — your OpenClaw session key\n task_label=\"my-task\", # Label shown in notifications\n chat_mode=\"agent\",\n)\n```\n\n### Wait for Completion (Universal)\n\nBlocks until CellCog finishes. Works with any agent — OpenClaw, Cursor, Claude Code, or any Python environment.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n task_label=\"my-task\",\n chat_mode=\"agent\",\n timeout=1800, # 30 min (default). Use 3600 for complex jobs.\n)\nprint(result[\"message\"])\nprint(result[\"status\"]) # \"completed\" | \"operating\" (+ timed_out=True if the wait gave up)\n```\n\n### When to Use Which\n\n| Scenario | Best Mode | Why |\n|----------|-----------|-----|\n| OpenClaw + long task + stay free | **Notify** | Agent keeps working, gets notified when done |\n| OpenClaw + chaining steps (research → summarize → PDF) | **Wait** | Each step feeds the next — simpler sequential workflows |\n| OpenClaw + quick task | **Either** | Both return fast for simple tasks |\n| Non-OpenClaw agent | **Wait** | Notify mode is OpenClaw-only |\n| Orchestrating SEVERAL chats in parallel | **Send-only** | Fire all chats at once, then poll — the blocking default would hang on each chat's whole run |\n\n**Notify mode** is more productive (agent never blocks).\n**Wait mode** is simpler to reason about, but blocks your agent for the duration.\n**Send-only** (`delivery=\"send_only\"`, works on `create_chat` AND `send_message`) returns the moment the chat/message is accepted — poll with `get_status()` (its `latest_update` field carries the agent's most recent progress line) and fetch results with `wait_for_completion()` or `get_history()`.\n\n### Continuing a Conversation\n\n```python\n# Wait mode (default)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n)\n\n# Notify mode (OpenClaw)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n notify_session_key=\"agent:main:main\",\n task_label=\"continue-research\",\n)\n```\n\n### Resuming After Timeout\n\nIf `create_chat()` or `wait_for_completion()` times out, CellCog is still working. The timeout response includes recent progress:\n\n```python\ncompletion = client.wait_for_completion(chat_id=\"abc123\", timeout=1800)\n```\n\n### Optional Parameters\n\n```python\nresult = client.create_chat(\n prompt=\"...\",\n task_label=\"...\",\n chat_mode=\"agent\", # See Chat Modes & Tiers below\n chat_tier=\"max\", # \"flash\" | \"core\" | \"max\" — omit for the SDK default\n delivery=\"send_only\", # fire-and-forget (see below); default blocks until done\n project_id=\"...\", # install project-management-cellcog for details\n agent_role_id=\"...\", # install project-management-cellcog for details\n enable_cowork=True, # install pair-programming-cellcog for details\n cowork_working_directory=\"/Users/...\", # install pair-programming-cellcog for details\n enable_browse=True, # drive the user's REAL Chrome (Desktop + extension required)\n browser_profile_id=\"Default\", # from client.get_browser_status()\n enable_tools=True, # the user's connected SaaS tools (Gmail, Notion, ...)\n tools_selection=[\"gmail\", \"notion\"], # toolkit slugs (omit for ALL); from client.list_toolkits()\n)\n```\n\n### Enabling Browse & Tools\n\nDiscover what's available on the user's account, then enable at chat creation:\n\n```python\n# Browse — discover Chrome profiles, enable with one\nstatus = client.get_browser_status()\n# status[\"available_profiles\"]: [{\"profileDir\", \"profileName\", \"hasExtension\", \"connected\"}, ...]\nresult = client.create_chat(\n prompt=\"Open the analytics dashboard and screenshot this week's numbers\",\n enable_browse=True,\n browser_profile_id=status[\"active_profile\"][\"profileDir\"],\n)\n\n# Tools — discover connected toolkits, enable a selection (omit tools_selection for ALL)\ntoolkits = client.list_toolkits(connected_only=True) # [{\"slug\": \"gmail\", ...}, ...]\nresult = client.create_chat(\n prompt=\"Summarize this week's unread emails\",\n enable_tools=True,\n tools_selection=[\"gmail\"], # TOOLKIT slugs; list_toolkit_tools(\"gmail\") shows what it unlocks\n)\n```\n\nBrowse requires CellCog Desktop + the Chrome extension and auto-enables co-work\nserver-side. If Browse isn't available, `create_chat` fails fast with a clear error\nbefore any credits are spent.\n\n---\n\n## Response Shape\n\nEvery SDK method returns the same shape:\n\n```python\n{\n \"chat_id\": str, # CellCog chat ID\n \"is_operating\": bool, # True = still working, False = done\n \"status\": str, # \"completed\" | \"tracking\" | \"accepted\" | \"operating\"\n # (a wait that gave up returns status=\"operating\" + timed_out=True —\n # the chat is STILL running server-side; resume with wait_for_completion)\n \"message\": str, # THE printable message — always print this in full\n}\n```\n\n**⚠️ Always print the entire `result[\"message\"]`.** Truncating or summarizing it will lose critical information including generated file paths, credits used, and follow-up instructions.\n\n### Utility Methods\n\n**`get_history(chat_id)`** — Full chat history (when original delivery was missed or you need to review). Returns the same shape; if still operating, `message` shows progress so far.\n\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n\n**`get_status(chat_id)`** — Lightweight status check (no history fetch):\n\n```python\nstatus = client.get_status(chat_id=\"abc123\")\nprint(status[\"is_operating\"]) # True/False\n```\n\n---\n\n## Chat Modes & Tiers\n\nEvery CellCog chat runs at a (mode, tier) operating point: `chat_mode` picks the agent, `chat_tier` picks the depth/spend.\n\n| Mode | Best For | Tiers | Min Credits |\n|------|----------|-------|-------------|\n| `\"agent\"` | Most tasks — assets, documents, coding, full production pipelines | `\"flash\"` / `\"core\"` / `\"max\"` | 60 |\n| `\"creative\"` | Design-taste work — dashboards, UI, brand identity, writing, slides | `\"core\"` / `\"max\"` (no flash) | 60 |\n| `\"team\"` | Deep research ONLY — multi-source synthesis, cross-validation, citations | `\"flash\"` / `\"core\"` / `\"max\"` | 1,000 (max tier: 2,000) |\n\n**Picking a tier in agent mode:**\n- **Omit `chat_tier`** → the SDK defaults to `\"flash\"` — right for simple asset generation and light tasks (fast, economical).\n- **Coding / co-work → `\"max\"`.** The SDK applies `\"max\"` automatically when `enable_cowork=True`.\n- **Heavy multi-step production** (video generation, data analysis, financial models, legal drafting) → `\"max\"`.\n- Quality disappointing on flash? Re-run the same prompt with `chat_tier=\"max\"`.\n\n**Use `\"team\"` only for deep research.** Agent max is now strong enough for almost every other use case — including video generation, which historically needed team mode and no longer does.\n\nLegacy mode names — `\"agent core\"`, `\"agent team\"`, `\"agent team max\"` — keep working forever (the server maps them to their historical operating points).\n\n---\n\n## Working with Files\n\n### Input: SHOW_FILE\n\nInclude local file paths in your prompt with `` tags (absolute paths required):\n\n```python\nprompt = \"\"\"\nAnalyze this sales data and create a report:\n/path/to/sales.csv\n\"\"\"\n```\n\n### Output: GENERATE_FILE\n\nUse `` tags to specify where output files should be stored on your machine. Essential for deterministic workflows where the next step needs to know the file path in advance.\n\n```python\nprompt = \"\"\"\nCreate a PDF report on Q4 earnings:\n/workspace/reports/q4_analysis.pdf\n\"\"\"\n```\n\nOutput downloads to the specified path instead of default `~/.cellcog/chats/{chat_id}/`.\n\n### File Downloads\n\nThe SDK automatically downloads files from CellCog responses:\n- **If you used `GENERATE_FILE` tags:** Files download to the path you specified\n- **Otherwise:** Files download to `~/.cellcog/chats/{chat_id}/`\n\nDownloaded file paths appear in `result[\"message\"]`. The SDK tracks seen messages — files are only downloaded once.\n\n**If you missed files or need to re-sync:**\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n`get_history()` re-processes the entire chat and downloads any missed files to their original destinations.\n\n---\n\n## Tips\n\n### ⚠️ CellCog Web Fallback\n\nEvery chat is accessible at https://cellcog.ai. When work gets complex or the SDK hits issues, direct your human to the web platform to view, continue, or take over directly.\n\n---\n\n## What CellCog Can Do\n\nCellCog is a sub-agent — not an API. Your agent offloads complex work to CellCog, which reasons, plans, and executes multi-tool workflows internally. A proprietary agent-to-agent communication protocol ensures high accuracy on first output, and because these are agent threads (not stateless API calls), every aspect of every generation can be refined through multi-step iteration.\n\nUnder the hood: frontier models across every domain, upgraded weekly. CellCog routes to the right models automatically — your agent just describes what it needs.\n\nInstall capability skills for detailed guidance:\n\n| Category | Skills |\n|----------|--------|\n| **Research & Analysis** | `deep-research-cellcog` `stock-analysis-cellcog` `crypto-research-cellcog` `data-analysis-cellcog` `news-briefing-cellcog` |\n| **Video & Cinema** | `video-generation-cellcog` `cinematic-video-cellcog` `instagram-reels-tiktok-cellcog` `youtube-video-cellcog` `seedance-video-generation-cellcog` |\n| **Images & Design** | `image-generation-cellcog` `logo-brand-identity-cellcog` `meme-generator-cellcog` `nano-banana-image-cellcog` `3d-model-generation-cellcog` `gif-generator-cellcog` `sticker-generator-cellcog` |\n| **Audio & Music** | `audio-generation-cellcog` `music-generation-cellcog` `podcast-generation-cellcog` |\n| **Avatars & Personas** | `avatar-creation-cellcog` |\n| **Documents & Slides** | `pdf-document-generation-cellcog` `presentation-slides-cellcog` `excel-spreadsheet-cellcog` `resume-cover-letter-cellcog` `legal-documents-cellcog` |\n| **Apps & Prototypes** | `dashboard-web-app-cellcog` `game-asset-generation-cellcog` `ui-prototype-wireframe-cellcog` `diagram-flowchart-cellcog` |\n| **Creative** | `comic-manga-generator-cellcog` `creative-writing-cellcog` `tutoring-education-cellcog` `travel-planning-cellcog` |\n| **Development** | `coding-agent-cellcog` `pair-programming-cellcog` `project-management-cellcog` `brainstorming-strategy-cellcog` |\n\n**This skill shows you HOW to use CellCog. Capability skills show you WHAT's possible.**\n\n---\n\n## OpenClaw Reference\n\n### Session Keys\n\nThe `notify_session_key` tells CellCog where to deliver results:\n\n| Context | Session Key |\n|---------|-------------|\n| Main agent | `\"agent:main:main\"` |\n| Sub-agent | `\"agent:main:subagent:{uuid}\"` |\n| Telegram DM | `\"agent:main:telegram:dm:{id}\"` |\n| Discord group | `\"agent:main:discord:group:{id}\"` |\n\n**Resilient delivery:** If your session ends before completion, results are automatically delivered to the parent session (e.g., sub-agent → main agent).\n\n### Sending Messages During Processing\n\nIn notify mode, your agent is free — you can send additional instructions to an operating chat at any time:\n\n```python\nclient.send_message(chat_id=\"abc123\", message=\"Actually focus only on Q4 data\",\n notify_session_key=\"agent:main:main\", task_label=\"refine\")\n\nclient.send_message(chat_id=\"abc123\", message=\"Stop operation\",\n notify_session_key=\"agent:main:main\", task_label=\"cancel\")\n```\n\nIn wait mode, your agent is blocked and cannot send messages until the current call returns.\n\n---\n\n## Support & Troubleshooting\n\nFor error handling, recovery patterns, ticket submission, and daemon troubleshooting:\n\n```python\ndocs = client.get_support_docs()\n```\n\nFile v2.0.21:_meta.json\n\n{\n \"ownerId\": \"kn7a96cj9q65e0bhmzahv790en80ffqm\",\n \"slug\": \"cellcog\",\n \"version\": \"2.0.21\",\n \"publishedAt\": 1787536261035\n}\n\nFile v2.0.21:skill-card.md\n\n## Description:\n\nCellCog lets agents send multimodal tasks to the CellCog service for research, analysis, generation, and deliverables such as images, video, audio, documents, dashboards, 3D models, diagrams, and code.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[cellcog](https://clawhub.ai/user/cellcog)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nDevelopers and agent users use this skill to offload multimodal research, analysis, generation, and coding tasks to CellCog, including requests that produce files or other deliverables.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: Files explicitly tagged for CellCog are uploaded to the CellCog service.\n\nMitigation: Only tag files intended for the task, and avoid sharing credentials, private keys, .env files, SSH keys, or other sensitive material.\n\nRisk: Optional browser and SaaS tool access can expose account or workspace data.\n\nMitigation: Enable browser or connected-tool access only when the task requires it, and choose the narrowest useful browser profile or toolkit selection.\n\nRisk: Using the CellCog SDK adds a locally installed dependency to the agent environment.\n\nMitigation: Install it only in an environment where running the SDK is acceptable, and consider pinning the package version.\n\n## Reference(s):\n\n- [ClawHub Skill Listing](https://clawhub.ai/cellcog/skills/cellcog)\n- [CellCog Homepage](https://cellcog.ai)\n- [CellCog Python SDK](https://github.com/CellCog/cellcog_python)\n- [CellCog PyPI Package](https://pypi.org/project/cellcog/)\n- [DeepResearch Bench Leaderboard](https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard)\n\n## Skill Output:\n\n**Output Type(s):** [Text, Markdown, Code, Shell commands, Configuration, Guidance, Files]\n\n**Output Format:** [Markdown guidance with Python and shell command examples, plus generated file paths when CellCog returns artifacts.]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [Responses may include downloaded artifact paths, completion status, and credit usage from the CellCog SDK.]\n\n## Skill Version(s):\n\n2.0.21 (source: release evidence)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.\n\nArchive v2.0.20: 3 files, 8905 bytes\n\nFiles: skill-card.md (2398b), SKILL.md (18222b), _meta.json (127b)\n\nFile v2.0.20:SKILL.md\n\n---\nname: cellcog\ndescription: \"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code.\"\nauthor: CellCog\nhomepage: https://cellcog.ai\nmetadata:\n openclaw:\n emoji: \"🧠\"\n os: [darwin, linux, windows]\n requires:\n bins: [python3]\n env: [CELLCOG_API_KEY]\n\n---\n# CellCog - Any-to-Any for Agents\n\n## The Power of Any-to-Any\n\nCellCog is the only AI that truly handles **any input → any output** in a single request. No tool chaining. No orchestration complexity. One call, multiple deliverables.\n\nCellCog pairs all modalities with frontier-level deep reasoning — as of July 2026, CellCog is **#1 on the DeepResearch Bench** (rankings change frequently; see the live leaderboard for the latest): https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard\n\n### Work With Multiple Files, Any Format\n\nReference as many documents as you need—all at once:\n\n```python\nprompt = \"\"\"\nAnalyze all of these together:\n/data/q4_earnings.pdf\n/data/competitor_analysis.pdf\n/data/market_research.xlsx\n/recordings/customer_interview.mp3\n/designs/product_mockup.png\n\nGive me a comprehensive market positioning analysis based on all these inputs.\n\"\"\"\n```\n\nFile paths must be absolute and enclosed in `` tags. CellCog understands PDFs, spreadsheets, images, audio, video, code files, and more.\n\n⚠️ **Without SHOW_FILE tags, CellCog only sees the path as text — not the file contents.**\n\n❌ `Analyze /data/sales.csv` — CellCog can't read the file\n✅ `Analyze /data/sales.csv` — CellCog reads it\n\n### Think of SHOW_FILE like reference files\n\nJust like Nano Banana accepts reference images, CellCog accepts reference files of any type — PDFs, spreadsheets, audio, code, images — as inputs the model reads during a task. Same mental model as any multimodal AI attachment.\n\n**Only attach what you intend to share.** Anything inside a `` tag is uploaded to CellCog. Don't wrap credentials, private keys, `.env` files, SSH keys, or other sensitive material in SHOW_FILE tags — the same way you wouldn't paste them into Nano Banana, ChatGPT, or any other AI service's file upload.\n\n### Request Multiple Outputs, Different Modalities\n\nAsk for completely different output types in ONE request:\n\n```python\nprompt = \"\"\"\nBased on this quarterly sales data:\n/data/sales_q4_2025.csv\n\nCreate ALL of the following:\n1. A PDF executive summary report with charts\n2. An interactive HTML dashboard for the leadership team\n3. A 60-second video presentation for the all-hands meeting\n4. A slide deck for the board presentation\n5. An Excel file with the underlying analysis and projections\n\"\"\"\n```\n\nCellCog handles the entire workflow — analyzing, generating, and delivering all outputs with consistent insights across every format.\n\n⚠️ **Be explicit about output artifacts.** Without explicit artifact language, CellCog may respond with text analysis instead of generating a file.\n\n❌ `\"Quarterly earnings analysis for AAPL\"` — could produce text or any format\n✅ `\"Create a PDF report and an interactive HTML dashboard analyzing AAPL quarterly earnings.\"` — CellCog creates actual deliverables\n\n**Your sub-agent for quality work.** Depth, accuracy, and real deliverables.\n\n---\n\n## Quick Start\n\n### Setup\n\n```python\nfrom cellcog import CellCogClient\n```\n\nIf import fails, install the official CellCog Python SDK:\n```bash\npip install -U cellcog\n```\n\n`cellcog` is the official Python SDK maintained by CellCog AI Inc. Source: https://github.com/CellCog/cellcog_python · Package: https://pypi.org/project/cellcog/\n\n### Authentication\n\n**Environment variable (recommended):** Set `CELLCOG_API_KEY` — the SDK picks it up automatically:\n```bash\nexport CELLCOG_API_KEY=\"sk_...\"\n```\n\nGet API key from: https://cellcog.ai/profile?tab=api-keys\n\n```python\nstatus = client.get_account_status()\nprint(status) # {\"configured\": True, \"email\": \"user@example.com\", ...}\n```\n\n### Agent Provider\n\n`agent_provider` is **required** when creating a `CellCogClient`. It identifies which agent framework is calling CellCog — not your individual agent's name, but the platform/tool you're running inside.\n\nExamples: `\"openclaw\"`, `\"claude-code\"`, `\"cursor\"`, `\"aider\"`, `\"windsurf\"`, `\"perplexity\"`, `\"hermes\"`, `\"script\"` (for standalone scripts).\n\n### OpenClaw Agents\n\nFire-and-forget — your agent stays free while CellCog works:\n\n```python\nclient = CellCogClient(agent_provider=\"openclaw\")\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n notify_session_key=\"agent:main:main\", # OpenClaw session key\n task_label=\"quantum-research\", # Label for notifications\n chat_mode=\"agent\",\n)\n# Returns IMMEDIATELY — daemon delivers results to your session when done\n```\n\n### All Other Agents (Cursor, Claude Code, etc.)\n\nBlocks until done — simplest pattern:\n\n```python\nclient = CellCogClient(agent_provider=\"cursor\") # or \"claude-code\", \"aider\", \"script\", etc.\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n task_label=\"quantum-research\",\n chat_mode=\"agent\",\n)\n# Blocks until done — result contains everything\nprint(result[\"message\"])\n```\n\n### Credit Usage\n\nCellCog orchestrates 21+ frontier foundation models. Credit consumption is unpredictable and varies by task complexity. Credits used are reported in every completion notification.\n\n---\n\n## Creating Tasks\n\n### Notify on Completion (OpenClaw — Fire-and-Forget)\n\nReturns immediately. A background daemon monitors via WebSocket and delivers results to your session when done. Your agent stays free to take new instructions, start other tasks, or continue working.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n notify_session_key=\"agent:main:main\", # Required — your OpenClaw session key\n task_label=\"my-task\", # Label shown in notifications\n chat_mode=\"agent\",\n)\n```\n\n### Wait for Completion (Universal)\n\nBlocks until CellCog finishes. Works with any agent — OpenClaw, Cursor, Claude Code, or any Python environment.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n task_label=\"my-task\",\n chat_mode=\"agent\",\n timeout=1800, # 30 min (default). Use 3600 for complex jobs.\n)\nprint(result[\"message\"])\nprint(result[\"status\"]) # \"completed\" | \"operating\" (+ timed_out=True if the wait gave up)\n```\n\n### When to Use Which\n\n| Scenario | Best Mode | Why |\n|----------|-----------|-----|\n| OpenClaw + long task + stay free | **Notify** | Agent keeps working, gets notified when done |\n| OpenClaw + chaining steps (research → summarize → PDF) | **Wait** | Each step feeds the next — simpler sequential workflows |\n| OpenClaw + quick task | **Either** | Both return fast for simple tasks |\n| Non-OpenClaw agent | **Wait** | Notify mode is OpenClaw-only |\n| Orchestrating SEVERAL chats in parallel | **Send-only** | Fire all chats at once, then poll — the blocking default would hang on each chat's whole run |\n\n**Notify mode** is more productive (agent never blocks).\n**Wait mode** is simpler to reason about, but blocks your agent for the duration.\n**Send-only** (`delivery=\"send_only\"`, works on `create_chat` AND `send_message`) returns the moment the chat/message is accepted — poll with `get_status()` (its `latest_update` field carries the agent's most recent progress line) and fetch results with `wait_for_completion()` or `get_history()`.\n\n### Continuing a Conversation\n\n```python\n# Wait mode (default)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n)\n\n# Notify mode (OpenClaw)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n notify_session_key=\"agent:main:main\",\n task_label=\"continue-research\",\n)\n```\n\n### Resuming After Timeout\n\nIf `create_chat()` or `wait_for_completion()` times out, CellCog is still working. The timeout response includes recent progress:\n\n```python\ncompletion = client.wait_for_completion(chat_id=\"abc123\", timeout=1800)\n```\n\n### Optional Parameters\n\n```python\nresult = client.create_chat(\n prompt=\"...\",\n task_label=\"...\",\n chat_mode=\"agent\", # See Chat Modes & Tiers below\n chat_tier=\"max\", # \"flash\" | \"core\" | \"max\" — omit for the SDK default\n delivery=\"send_only\", # fire-and-forget (see below); default blocks until done\n project_id=\"...\", # install project-management-cellcog for details\n agent_role_id=\"...\", # install project-management-cellcog for details\n enable_cowork=True, # install pair-programming-cellcog for details\n cowork_working_directory=\"/Users/...\", # install pair-programming-cellcog for details\n enable_browse=True, # drive the user's REAL Chrome (Desktop + extension required)\n browser_profile_id=\"Default\", # from client.get_browser_status()\n enable_tools=True, # the user's connected SaaS tools (Gmail, Notion, ...)\n tools_selection=[\"gmail\", \"notion\"], # toolkit slugs (omit for ALL); from client.list_toolkits()\n)\n```\n\n### Enabling Browse & Tools\n\nDiscover what's available on the user's account, then enable at chat creation:\n\n```python\n# Browse — discover Chrome profiles, enable with one\nstatus = client.get_browser_status()\n# status[\"available_profiles\"]: [{\"profileDir\", \"profileName\", \"hasExtension\", \"connected\"}, ...]\nresult = client.create_chat(\n prompt=\"Open the analytics dashboard and screenshot this week's numbers\",\n enable_browse=True,\n browser_profile_id=status[\"active_profile\"][\"profileDir\"],\n)\n\n# Tools — discover connected toolkits, enable a selection (omit tools_selection for ALL)\ntoolkits = client.list_toolkits(connected_only=True) # [{\"slug\": \"gmail\", ...}, ...]\nresult = client.create_chat(\n prompt=\"Summarize this week's unread emails\",\n enable_tools=True,\n tools_selection=[\"gmail\"], # TOOLKIT slugs; list_toolkit_tools(\"gmail\") shows what it unlocks\n)\n```\n\nBrowse requires CellCog Desktop + the Chrome extension and auto-enables co-work\nserver-side. If Browse isn't available, `create_chat` fails fast with a clear error\nbefore any credits are spent.\n\n---\n\n## Response Shape\n\nEvery SDK method returns the same shape:\n\n```python\n{\n \"chat_id\": str, # CellCog chat ID\n \"is_operating\": bool, # True = still working, False = done\n \"status\": str, # \"completed\" | \"tracking\" | \"accepted\" | \"operating\"\n # (a wait that gave up returns status=\"operating\" + timed_out=True —\n # the chat is STILL running server-side; resume with wait_for_completion)\n \"message\": str, # THE printable message — always print this in full\n}\n```\n\n**⚠️ Always print the entire `result[\"message\"]`.** Truncating or summarizing it will lose critical information including generated file paths, credits used, and follow-up instructions.\n\n### Utility Methods\n\n**`get_history(chat_id)`** — Full chat history (when original delivery was missed or you need to review). Returns the same shape; if still operating, `message` shows progress so far.\n\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n\n**`get_status(chat_id)`** — Lightweight status check (no history fetch):\n\n```python\nstatus = client.get_status(chat_id=\"abc123\")\nprint(status[\"is_operating\"]) # True/False\n```\n\n---\n\n## Chat Modes & Tiers\n\nEvery CellCog chat runs at a (mode, tier) operating point: `chat_mode` picks the agent, `chat_tier` picks the depth/spend.\n\n| Mode | Best For | Tiers | Min Credits |\n|------|----------|-------|-------------|\n| `\"agent\"` | Most tasks — assets, documents, coding, full production pipelines | `\"flash\"` / `\"core\"` / `\"max\"` | 60 |\n| `\"creative\"` | Design-taste work — dashboards, UI, brand identity, writing, slides | `\"core\"` / `\"max\"` (no flash) | 60 |\n| `\"team\"` | Deep research ONLY — multi-source synthesis, cross-validation, citations | `\"flash\"` / `\"core\"` / `\"max\"` | 1,000 (max tier: 2,000) |\n\n**Picking a tier in agent mode:**\n- **Omit `chat_tier`** → the SDK defaults to `\"flash\"` — right for simple asset generation and light tasks (fast, economical).\n- **Coding / co-work → `\"max\"`.** The SDK applies `\"max\"` automatically when `enable_cowork=True`.\n- **Heavy multi-step production** (video generation, data analysis, financial models, legal drafting) → `\"max\"`.\n- Quality disappointing on flash? Re-run the same prompt with `chat_tier=\"max\"`.\n\n**Use `\"team\"` only for deep research.** Agent max is now strong enough for almost every other use case — including video generation, which historically needed team mode and no longer does.\n\nLegacy mode names — `\"agent core\"`, `\"agent team\"`, `\"agent team max\"` — keep working forever (the server maps them to their historical operating points).\n\n---\n\n## Working with Files\n\n### Input: SHOW_FILE\n\nInclude local file paths in your prompt with `` tags (absolute paths required):\n\n```python\nprompt = \"\"\"\nAnalyze this sales data and create a report:\n/path/to/sales.csv\n\"\"\"\n```\n\n### Output: GENERATE_FILE\n\nUse `` tags to specify where output files should be stored on your machine. Essential for deterministic workflows where the next step needs to know the file path in advance.\n\n```python\nprompt = \"\"\"\nCreate a PDF report on Q4 earnings:\n/workspace/reports/q4_analysis.pdf\n\"\"\"\n```\n\nOutput downloads to the specified path instead of default `~/.cellcog/chats/{chat_id}/`.\n\n### File Downloads\n\nThe SDK automatically downloads files from CellCog responses:\n- **If you used `GENERATE_FILE` tags:** Files download to the path you specified\n- **Otherwise:** Files download to `~/.cellcog/chats/{chat_id}/`\n\nDownloaded file paths appear in `result[\"message\"]`. The SDK tracks seen messages — files are only downloaded once.\n\n**If you missed files or need to re-sync:**\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n`get_history()` re-processes the entire chat and downloads any missed files to their original destinations.\n\n---\n\n## Tips\n\n### ⚠️ CellCog Web Fallback\n\nEvery chat is accessible at https://cellcog.ai. When work gets complex or the SDK hits issues, direct your human to the web platform to view, continue, or take over directly.\n\n---\n\n## What CellCog Can Do\n\nCellCog is a sub-agent — not an API. Your agent offloads complex work to CellCog, which reasons, plans, and executes multi-tool workflows internally. A proprietary agent-to-agent communication protocol ensures high accuracy on first output, and because these are agent threads (not stateless API calls), every aspect of every generation can be refined through multi-step iteration.\n\nUnder the hood: frontier models across every domain, upgraded weekly. CellCog routes to the right models automatically — your agent just describes what it needs.\n\nInstall capability skills for detailed guidance:\n\n| Category | Skills |\n|----------|--------|\n| **Research & Analysis** | `deep-research-cellcog` `stock-analysis-cellcog` `crypto-research-cellcog` `data-analysis-cellcog` `news-briefing-cellcog` |\n| **Video & Cinema** | `video-generation-cellcog` `cinematic-video-cellcog` `instagram-reels-tiktok-cellcog` `youtube-video-cellcog` `seedance-video-generation-cellcog` |\n| **Images & Design** | `image-generation-cellcog` `logo-brand-identity-cellcog` `meme-generator-cellcog` `nano-banana-image-cellcog` `3d-model-generation-cellcog` `gif-generator-cellcog` `sticker-generator-cellcog` |\n| **Audio & Music** | `audio-generation-cellcog` `music-generation-cellcog` `podcast-generation-cellcog` |\n| **Avatars & Personas** | `avatar-creation-cellcog` |\n| **Documents & Slides** | `pdf-document-generation-cellcog` `presentation-slides-cellcog` `excel-spreadsheet-cellcog` `resume-cover-letter-cellcog` `legal-documents-cellcog` |\n| **Apps & Prototypes** | `dashboard-web-app-cellcog` `game-asset-generation-cellcog` `ui-prototype-wireframe-cellcog` `diagram-flowchart-cellcog` |\n| **Creative** | `comic-manga-generator-cellcog` `creative-writing-cellcog` `tutoring-education-cellcog` `travel-planning-cellcog` |\n| **Development** | `coding-agent-cellcog` `pair-programming-cellcog` `project-management-cellcog` `brainstorming-strategy-cellcog` |\n\n**This skill shows you HOW to use CellCog. Capability skills show you WHAT's possible.**\n\n---\n\n## OpenClaw Reference\n\n### Session Keys\n\nThe `notify_session_key` tells CellCog where to deliver results:\n\n| Context | Session Key |\n|---------|-------------|\n| Main agent | `\"agent:main:main\"` |\n| Sub-agent | `\"agent:main:subagent:{uuid}\"` |\n| Telegram DM | `\"agent:main:telegram:dm:{id}\"` |\n| Discord group | `\"agent:main:discord:group:{id}\"` |\n\n**Resilient delivery:** If your session ends before completion, results are automatically delivered to the parent session (e.g., sub-agent → main agent).\n\n### Sending Messages During Processing\n\nIn notify mode, your agent is free — you can send additional instructions to an operating chat at any time:\n\n```python\nclient.send_message(chat_id=\"abc123\", message=\"Actually focus only on Q4 data\",\n notify_session_key=\"agent:main:main\", task_label=\"refine\")\n\nclient.send_message(chat_id=\"abc123\", message=\"Stop operation\",\n notify_session_key=\"agent:main:main\", task_label=\"cancel\")\n```\n\nIn wait mode, your agent is blocked and cannot send messages until the current call returns.\n\n---\n\n## Support & Troubleshooting\n\nFor error handling, recovery patterns, ticket submission, and daemon troubleshooting:\n\n```python\ndocs = client.get_support_docs()\n```\n\nFile v2.0.20:_meta.json\n\n{\n \"ownerId\": \"kn7a96cj9q65e0bhmzahv790en80ffqm\",\n \"slug\": \"cellcog\",\n \"version\": \"2.0.20\",\n \"publishedAt\": 1787535914207\n}\n\nFile v2.0.20:skill-card.md\n\n## Description:\n\nCellCog helps agents offload multimodal tasks to the CellCog service to generate research, documents, media, code, and other deliverables from text and file inputs.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[cellcog](https://clawhub.ai/user/cellcog)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nDevelopers and external agent users use this skill to connect an agent to CellCog for multimodal research, analysis, content generation, file-based work, and generated deliverables.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: Local files wrapped in SHOW_FILE tags may be uploaded to the CellCog external AI service.\n\nMitigation: Only wrap files that are intended for upload, and exclude secrets, private keys, credentials, .env files, and other sensitive material.\n\nRisk: Connected browser or SaaS tools can expose account data within the enabled scopes.\n\nMitigation: Enable browser, Gmail, Notion, or similar tools only for accounts and scopes that are appropriate to share with CellCog.\n\nRisk: Generated deliverables or long-running task results may include important file paths, credit usage, and follow-up instructions.\n\nMitigation: Review and preserve the full returned message rather than truncating or summarizing it.\n\n## Reference(s):\n\n- [ClawHub Skill Page](https://clawhub.ai/cellcog/skills/cellcog)\n- [CellCog Homepage](https://cellcog.ai)\n- [CellCog Python SDK Source](https://github.com/CellCog/cellcog_python)\n- [CellCog Python SDK Package](https://pypi.org/project/cellcog/)\n- [DeepResearch Bench Leaderboard](https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard)\n- [CellCog API Key Page](https://cellcog.ai/profile?tab=api-keys)\n\n## Skill Output:\n\n**Output Type(s):** [Guidance, Code, Shell commands, Configuration]\n\n**Output Format:** [Markdown with Python and bash code blocks]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [May guide the agent to upload tagged local files to CellCog and print returned messages in full.]\n\n## Skill Version(s):\n\n2.0.20 (source: server release metadata)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.\n\nArchive v2.0.19: 3 files, 7952 bytes\n\nFiles: skill-card.md (2488b), SKILL.md (15510b), _meta.json (127b)\n\nFile v2.0.19:SKILL.md\n\n---\nname: cellcog\ndescription: \"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code.\"\nauthor: CellCog\nhomepage: https://cellcog.ai\nmetadata:\n openclaw:\n emoji: \"🧠\"\n os: [darwin, linux, windows]\n requires:\n bins: [python3]\n env: [CELLCOG_API_KEY]\n\n---\n# CellCog - Any-to-Any for Agents\n\n## The Power of Any-to-Any\n\nCellCog is the only AI that truly handles **any input → any output** in a single request. No tool chaining. No orchestration complexity. One call, multiple deliverables.\n\nCellCog pairs all modalities with frontier-level deep reasoning — as of July 2026, CellCog is **#1 on the DeepResearch Bench** (rankings change frequently; see the live leaderboard for the latest): https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard\n\n### Work With Multiple Files, Any Format\n\nReference as many documents as you need—all at once:\n\n```python\nprompt = \"\"\"\nAnalyze all of these together:\n/data/q4_earnings.pdf\n/data/competitor_analysis.pdf\n/data/market_research.xlsx\n/recordings/customer_interview.mp3\n/designs/product_mockup.png\n\nGive me a comprehensive market positioning analysis based on all these inputs.\n\"\"\"\n```\n\nFile paths must be absolute and enclosed in `` tags. CellCog understands PDFs, spreadsheets, images, audio, video, code files, and more.\n\n⚠️ **Without SHOW_FILE tags, CellCog only sees the path as text — not the file contents.**\n\n❌ `Analyze /data/sales.csv` — CellCog can't read the file\n✅ `Analyze /data/sales.csv` — CellCog reads it\n\n### Think of SHOW_FILE like reference files\n\nJust like Nano Banana accepts reference images, CellCog accepts reference files of any type — PDFs, spreadsheets, audio, code, images — as inputs the model reads during a task. Same mental model as any multimodal AI attachment.\n\n**Only attach what you intend to share.** Anything inside a `` tag is uploaded to CellCog. Don't wrap credentials, private keys, `.env` files, SSH keys, or other sensitive material in SHOW_FILE tags — the same way you wouldn't paste them into Nano Banana, ChatGPT, or any other AI service's file upload.\n\n### Request Multiple Outputs, Different Modalities\n\nAsk for completely different output types in ONE request:\n\n```python\nprompt = \"\"\"\nBased on this quarterly sales data:\n/data/sales_q4_2025.csv\n\nCreate ALL of the following:\n1. A PDF executive summary report with charts\n2. An interactive HTML dashboard for the leadership team\n3. A 60-second video presentation for the all-hands meeting\n4. A slide deck for the board presentation\n5. An Excel file with the underlying analysis and projections\n\"\"\"\n```\n\nCellCog handles the entire workflow — analyzing, generating, and delivering all outputs with consistent insights across every format.\n\n⚠️ **Be explicit about output artifacts.** Without explicit artifact language, CellCog may respond with text analysis instead of generating a file.\n\n❌ `\"Quarterly earnings analysis for AAPL\"` — could produce text or any format\n✅ `\"Create a PDF report and an interactive HTML dashboard analyzing AAPL quarterly earnings.\"` — CellCog creates actual deliverables\n\n**Your sub-agent for quality work.** Depth, accuracy, and real deliverables.\n\n---\n\n## Quick Start\n\n### Setup\n\n```python\nfrom cellcog import CellCogClient\n```\n\nIf import fails, install the official CellCog Python SDK:\n```bash\npip install -U cellcog\n```\n\n`cellcog` is the official Python SDK maintained by CellCog AI Inc. Source: https://github.com/CellCog/cellcog_python · Package: https://pypi.org/project/cellcog/\n\n### Authentication\n\n**Environment variable (recommended):** Set `CELLCOG_API_KEY` — the SDK picks it up automatically:\n```bash\nexport CELLCOG_API_KEY=\"sk_...\"\n```\n\nGet API key from: https://cellcog.ai/profile?tab=api-keys\n\n```python\nstatus = client.get_account_status()\nprint(status) # {\"configured\": True, \"email\": \"user@example.com\", ...}\n```\n\n### Agent Provider\n\n`agent_provider` is **required** when creating a `CellCogClient`. It identifies which agent framework is calling CellCog — not your individual agent's name, but the platform/tool you're running inside.\n\nExamples: `\"openclaw\"`, `\"claude-code\"`, `\"cursor\"`, `\"aider\"`, `\"windsurf\"`, `\"perplexity\"`, `\"hermes\"`, `\"script\"` (for standalone scripts).\n\n### OpenClaw Agents\n\nFire-and-forget — your agent stays free while CellCog works:\n\n```python\nclient = CellCogClient(agent_provider=\"openclaw\")\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n notify_session_key=\"agent:main:main\", # OpenClaw session key\n task_label=\"quantum-research\", # Label for notifications\n chat_mode=\"agent\",\n)\n# Returns IMMEDIATELY — daemon delivers results to your session when done\n```\n\n### All Other Agents (Cursor, Claude Code, etc.)\n\nBlocks until done — simplest pattern:\n\n```python\nclient = CellCogClient(agent_provider=\"cursor\") # or \"claude-code\", \"aider\", \"script\", etc.\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n task_label=\"quantum-research\",\n chat_mode=\"agent\",\n)\n# Blocks until done — result contains everything\nprint(result[\"message\"])\n```\n\n### Credit Usage\n\nCellCog orchestrates 21+ frontier foundation models. Credit consumption is unpredictable and varies by task complexity. Credits used are reported in every completion notification.\n\n---\n\n## Creating Tasks\n\n### Notify on Completion (OpenClaw — Fire-and-Forget)\n\nReturns immediately. A background daemon monitors via WebSocket and delivers results to your session when done. Your agent stays free to take new instructions, start other tasks, or continue working.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n notify_session_key=\"agent:main:main\", # Required — your OpenClaw session key\n task_label=\"my-task\", # Label shown in notifications\n chat_mode=\"agent\",\n)\n```\n\n### Wait for Completion (Universal)\n\nBlocks until CellCog finishes. Works with any agent — OpenClaw, Cursor, Claude Code, or any Python environment.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n task_label=\"my-task\",\n chat_mode=\"agent\",\n timeout=1800, # 30 min (default). Use 3600 for complex jobs.\n)\nprint(result[\"message\"])\nprint(result[\"status\"]) # \"completed\" | \"timeout\"\n```\n\n### When to Use Which\n\n| Scenario | Best Mode | Why |\n|----------|-----------|-----|\n| OpenClaw + long task + stay free | **Notify** | Agent keeps working, gets notified when done |\n| OpenClaw + chaining steps (research → summarize → PDF) | **Wait** | Each step feeds the next — simpler sequential workflows |\n| OpenClaw + quick task | **Either** | Both return fast for simple tasks |\n| Non-OpenClaw agent | **Wait** | Notify mode is OpenClaw-only |\n\n**Notify mode** is more productive (agent never blocks).\n**Wait mode** is simpler to reason about, but blocks your agent for the duration.\n\n### Continuing a Conversation\n\n```python\n# Wait mode (default)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n)\n\n# Notify mode (OpenClaw)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n notify_session_key=\"agent:main:main\",\n task_label=\"continue-research\",\n)\n```\n\n### Resuming After Timeout\n\nIf `create_chat()` or `wait_for_completion()` times out, CellCog is still working. The timeout response includes recent progress:\n\n```python\ncompletion = client.wait_for_completion(chat_id=\"abc123\", timeout=1800)\n```\n\n### Optional Parameters\n\n```python\nresult = client.create_chat(\n prompt=\"...\",\n task_label=\"...\",\n chat_mode=\"agent\", # See Chat Modes below\n project_id=\"...\", # install project-management-cellcog for details\n agent_role_id=\"...\", # install project-management-cellcog for details\n enable_cowork=True, # install pair-programming-cellcog for details\n cowork_working_directory=\"/Users/...\", # install pair-programming-cellcog for details\n)\n```\n\n---\n\n## Response Shape\n\nEvery SDK method returns the same shape:\n\n```python\n{\n \"chat_id\": str, # CellCog chat ID\n \"is_operating\": bool, # True = still working, False = done\n \"status\": str, # \"completed\" | \"tracking\" | \"timeout\" | \"operating\"\n \"message\": str, # THE printable message — always print this in full\n}\n```\n\n**⚠️ Always print the entire `result[\"message\"]`.** Truncating or summarizing it will lose critical information including generated file paths, credits used, and follow-up instructions.\n\n### Utility Methods\n\n**`get_history(chat_id)`** — Full chat history (when original delivery was missed or you need to review). Returns the same shape; if still operating, `message` shows progress so far.\n\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n\n**`get_status(chat_id)`** — Lightweight status check (no history fetch):\n\n```python\nstatus = client.get_status(chat_id=\"abc123\")\nprint(status[\"is_operating\"]) # True/False\n```\n\n---\n\n## Chat Modes\n\n| Mode | Best For | Speed | Min Credits |\n|------|----------|-------|-------------|\n| `\"agent\"` | Most tasks — images, audio, dashboards, spreadsheets, presentations | Fast (seconds to minutes) | 100 |\n| `\"agent core\"` | Coding, co-work, terminal operations | Fast | 50 |\n| `\"agent team\"` | Deep research & multi-angled reasoning across every modality | Slower (5-60 min) | 500 |\n| `\"agent team max\"` | High-stakes work where extra reasoning depth justifies the cost | Slowest | 2,000 |\n\n- **`\"agent\"` (default)** — Most versatile. Handles most tasks excellently, including deep research when guided.\n- **`\"agent core\"`** — Lightweight context for code, terminal, and file operations. Multimedia tools load on demand. Requires Co-work (CellCog Desktop). See `coding-agent-cellcog`.\n- **`\"agent team\"`** — A team of agents that debates, cross-validates, and delivers comprehensive results. The only platform with deep reasoning across every modality.\n- **`\"agent team max\"`** — Same Agent Team with all settings maxed. Quality gain is incremental (5-10%) but meaningful for costly decisions.\n\n---\n\n## Working with Files\n\n### Input: SHOW_FILE\n\nInclude local file paths in your prompt with `` tags (absolute paths required):\n\n```python\nprompt = \"\"\"\nAnalyze this sales data and create a report:\n/path/to/sales.csv\n\"\"\"\n```\n\n### Output: GENERATE_FILE\n\nUse `` tags to specify where output files should be stored on your machine. Essential for deterministic workflows where the next step needs to know the file path in advance.\n\n```python\nprompt = \"\"\"\nCreate a PDF report on Q4 earnings:\n/workspace/reports/q4_analysis.pdf\n\"\"\"\n```\n\nOutput downloads to the specified path instead of default `~/.cellcog/chats/{chat_id}/`.\n\n### File Downloads\n\nThe SDK automatically downloads files from CellCog responses:\n- **If you used `GENERATE_FILE` tags:** Files download to the path you specified\n- **Otherwise:** Files download to `~/.cellcog/chats/{chat_id}/`\n\nDownloaded file paths appear in `result[\"message\"]`. The SDK tracks seen messages — files are only downloaded once.\n\n**If you missed files or need to re-sync:**\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n`get_history()` re-processes the entire chat and downloads any missed files to their original destinations.\n\n---\n\n## Tips\n\n### ⚠️ CellCog Web Fallback\n\nEvery chat is accessible at https://cellcog.ai. When work gets complex or the SDK hits issues, direct your human to the web platform to view, continue, or take over directly.\n\n---\n\n## What CellCog Can Do\n\nCellCog is a sub-agent — not an API. Your agent offloads complex work to CellCog, which reasons, plans, and executes multi-tool workflows internally. A proprietary agent-to-agent communication protocol ensures high accuracy on first output, and because these are agent threads (not stateless API calls), every aspect of every generation can be refined through multi-step iteration.\n\nUnder the hood: frontier models across every domain, upgraded weekly. CellCog routes to the right models automatically — your agent just describes what it needs.\n\nInstall capability skills for detailed guidance:\n\n| Category | Skills |\n|----------|--------|\n| **Research & Analysis** | `deep-research-cellcog` `stock-analysis-cellcog` `crypto-research-cellcog` `data-analysis-cellcog` `news-briefing-cellcog` |\n| **Video & Cinema** | `video-generation-cellcog` `cinematic-video-cellcog` `instagram-reels-tiktok-cellcog` `youtube-video-cellcog` `seedance-video-generation-cellcog` |\n| **Images & Design** | `image-generation-cellcog` `logo-brand-identity-cellcog` `meme-generator-cellcog` `nano-banana-image-cellcog` `3d-model-generation-cellcog` `gif-generator-cellcog` `sticker-generator-cellcog` |\n| **Audio & Music** | `audio-generation-cellcog` `music-generation-cellcog` `podcast-generation-cellcog` |\n| **Avatars & Personas** | `avatar-creation-cellcog` |\n| **Documents & Slides** | `pdf-document-generation-cellcog` `presentation-slides-cellcog` `excel-spreadsheet-cellcog` `resume-cover-letter-cellcog` `legal-documents-cellcog` |\n| **Apps & Prototypes** | `dashboard-web-app-cellcog` `game-asset-generation-cellcog` `ui-prototype-wireframe-cellcog` `diagram-flowchart-cellcog` |\n| **Creative** | `comic-manga-generator-cellcog` `creative-writing-cellcog` `tutoring-education-cellcog` `travel-planning-cellcog` |\n| **Development** | `coding-agent-cellcog` `pair-programming-cellcog` `project-management-cellcog` `brainstorming-strategy-cellcog` |\n\n**This skill shows you HOW to use CellCog. Capability skills show you WHAT's possible.**\n\n---\n\n## OpenClaw Reference\n\n### Session Keys\n\nThe `notify_session_key` tells CellCog where to deliver results:\n\n| Context | Session Key |\n|---------|-------------|\n| Main agent | `\"agent:main:main\"` |\n| Sub-agent | `\"agent:main:subagent:{uuid}\"` |\n| Telegram DM | `\"agent:main:telegram:dm:{id}\"` |\n| Discord group | `\"agent:main:discord:group:{id}\"` |\n\n**Resilient delivery:** If your session ends before completion, results are automatically delivered to the parent session (e.g., sub-agent → main agent).\n\n### Sending Messages During Processing\n\nIn notify mode, your agent is free — you can send additional instructions to an operating chat at any time:\n\n```python\nclient.send_message(chat_id=\"abc123\", message=\"Actually focus only on Q4 data\",\n notify_session_key=\"agent:main:main\", task_label=\"refine\")\n\nclient.send_message(chat_id=\"abc123\", message=\"Stop operation\",\n notify_session_key=\"agent:main:main\", task_label=\"cancel\")\n```\n\nIn wait mode, your agent is blocked and cannot send messages until the current call returns.\n\n---\n\n## Support & Troubleshooting\n\nFor error handling, recovery patterns, ticket submission, and daemon troubleshooting:\n\n```python\ndocs = client.get_support_docs()\n```\n\nFile v2.0.19:_meta.json\n\n{\n \"ownerId\": \"kn7a96cj9q65e0bhmzahv790en80ffqm\",\n \"slug\": \"cellcog\",\n \"version\": \"2.0.19\",\n \"publishedAt\": 1786288771179\n}\n\nFile v2.0.19:skill-card.md\n\n## Description:\n\nAny-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[cellcog](https://clawhub.ai/user/cellcog)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nDevelopers and agent users use this skill to delegate multimodal research, analysis, generation, coding, and document tasks to the CellCog remote AI sub-agent. It provides setup, authentication, file-sharing, task creation, and result-handling guidance for CellCog SDK workflows.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: Prompts and files explicitly wrapped in SHOW_FILE tags are sent to CellCog.\n\nMitigation: Share only files intended for CellCog processing, and exclude secrets, private keys, .env files, credentials, and other sensitive material.\n\nRisk: Full result messages may expose generated file paths, credit usage, or follow-up details in shared or logged environments.\n\nMitigation: Review full result messages in an appropriate environment and redact sensitive details before sharing logs or transcripts.\n\n## Reference(s):\n\n- [CellCog skill page](https://clawhub.ai/cellcog/skills/cellcog)\n- [CellCog homepage](https://cellcog.ai)\n- [CellCog Python SDK](https://github.com/CellCog/cellcog_python)\n- [cellcog Python package](https://pypi.org/project/cellcog/)\n- [DeepResearch Bench Leaderboard](https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard)\n\n## Skill Output:\n\n**Output Type(s):** [text, markdown, code, shell commands, configuration, guidance, files]\n\n**Output Format:** [Markdown guidance with Python and shell code examples]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [May describe generated artifacts, downloaded file paths, credit usage, status, and follow-up instructions returned by CellCog.]\n\n## Skill Version(s):\n\n2.0.19 (source: server release evidence)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.\n\nArchive v2.0.18: 3 files, 7899 bytes\n\nFiles: skill-card.md (2502b), SKILL.md (15510b), _meta.json (127b)\n\nFile v2.0.18:SKILL.md\n\n---\nname: cellcog\ndescription: \"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code.\"\nauthor: CellCog\nhomepage: https://cellcog.ai\nmetadata:\n openclaw:\n emoji: \"🧠\"\n os: [darwin, linux, windows]\n requires:\n bins: [python3]\n env: [CELLCOG_API_KEY]\n\n---\n# CellCog - Any-to-Any for Agents\n\n## The Power of Any-to-Any\n\nCellCog is the only AI that truly handles **any input → any output** in a single request. No tool chaining. No orchestration complexity. One call, multiple deliverables.\n\nCellCog pairs all modalities with frontier-level deep reasoning — as of July 2026, CellCog is **#1 on the DeepResearch Bench** (rankings change frequently; see the live leaderboard for the latest): https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard\n\n### Work With Multiple Files, Any Format\n\nReference as many documents as you need—all at once:\n\n```python\nprompt = \"\"\"\nAnalyze all of these together:\n/data/q4_earnings.pdf\n/data/competitor_analysis.pdf\n/data/market_research.xlsx\n/recordings/customer_interview.mp3\n/designs/product_mockup.png\n\nGive me a comprehensive market positioning analysis based on all these inputs.\n\"\"\"\n```\n\nFile paths must be absolute and enclosed in `` tags. CellCog understands PDFs, spreadsheets, images, audio, video, code files, and more.\n\n⚠️ **Without SHOW_FILE tags, CellCog only sees the path as text — not the file contents.**\n\n❌ `Analyze /data/sales.csv` — CellCog can't read the file\n✅ `Analyze /data/sales.csv` — CellCog reads it\n\n### Think of SHOW_FILE like reference files\n\nJust like Nano Banana accepts reference images, CellCog accepts reference files of any type — PDFs, spreadsheets, audio, code, images — as inputs the model reads during a task. Same mental model as any multimodal AI attachment.\n\n**Only attach what you intend to share.** Anything inside a `` tag is uploaded to CellCog. Don't wrap credentials, private keys, `.env` files, SSH keys, or other sensitive material in SHOW_FILE tags — the same way you wouldn't paste them into Nano Banana, ChatGPT, or any other AI service's file upload.\n\n### Request Multiple Outputs, Different Modalities\n\nAsk for completely different output types in ONE request:\n\n```python\nprompt = \"\"\"\nBased on this quarterly sales data:\n/data/sales_q4_2025.csv\n\nCreate ALL of the following:\n1. A PDF executive summary report with charts\n2. An interactive HTML dashboard for the leadership team\n3. A 60-second video presentation for the all-hands meeting\n4. A slide deck for the board presentation\n5. An Excel file with the underlying analysis and projections\n\"\"\"\n```\n\nCellCog handles the entire workflow — analyzing, generating, and delivering all outputs with consistent insights across every format.\n\n⚠️ **Be explicit about output artifacts.** Without explicit artifact language, CellCog may respond with text analysis instead of generating a file.\n\n❌ `\"Quarterly earnings analysis for AAPL\"` — could produce text or any format\n✅ `\"Create a PDF report and an interactive HTML dashboard analyzing AAPL quarterly earnings.\"` — CellCog creates actual deliverables\n\n**Your sub-agent for quality work.** Depth, accuracy, and real deliverables.\n\n---\n\n## Quick Start\n\n### Setup\n\n```python\nfrom cellcog import CellCogClient\n```\n\nIf import fails, install the official CellCog Python SDK:\n```bash\npip install -U cellcog\n```\n\n`cellcog` is the official Python SDK maintained by CellCog AI Inc. Source: https://github.com/CellCog/cellcog_python · Package: https://pypi.org/project/cellcog/\n\n### Authentication\n\n**Environment variable (recommended):** Set `CELLCOG_API_KEY` — the SDK picks it up automatically:\n```bash\nexport CELLCOG_API_KEY=\"sk_...\"\n```\n\nGet API key from: https://cellcog.ai/profile?tab=api-keys\n\n```python\nstatus = client.get_account_status()\nprint(status) # {\"configured\": True, \"email\": \"user@example.com\", ...}\n```\n\n### Agent Provider\n\n`agent_provider` is **required** when creating a `CellCogClient`. It identifies which agent framework is calling CellCog — not your individual agent's name, but the platform/tool you're running inside.\n\nExamples: `\"openclaw\"`, `\"claude-code\"`, `\"cursor\"`, `\"aider\"`, `\"windsurf\"`, `\"perplexity\"`, `\"hermes\"`, `\"script\"` (for standalone scripts).\n\n### OpenClaw Agents\n\nFire-and-forget — your agent stays free while CellCog works:\n\n```python\nclient = CellCogClient(agent_provider=\"openclaw\")\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n notify_session_key=\"agent:main:main\", # OpenClaw session key\n task_label=\"quantum-research\", # Label for notifications\n chat_mode=\"agent\",\n)\n# Returns IMMEDIATELY — daemon delivers results to your session when done\n```\n\n### All Other Agents (Cursor, Claude Code, etc.)\n\nBlocks until done — simplest pattern:\n\n```python\nclient = CellCogClient(agent_provider=\"cursor\") # or \"claude-code\", \"aider\", \"script\", etc.\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n task_label=\"quantum-research\",\n chat_mode=\"agent\",\n)\n# Blocks until done — result contains everything\nprint(result[\"message\"])\n```\n\n### Credit Usage\n\nCellCog orchestrates 21+ frontier foundation models. Credit consumption is unpredictable and varies by task complexity. Credits used are reported in every completion notification.\n\n---\n\n## Creating Tasks\n\n### Notify on Completion (OpenClaw — Fire-and-Forget)\n\nReturns immediately. A background daemon monitors via WebSocket and delivers results to your session when done. Your agent stays free to take new instructions, start other tasks, or continue working.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n notify_session_key=\"agent:main:main\", # Required — your OpenClaw session key\n task_label=\"my-task\", # Label shown in notifications\n chat_mode=\"agent\",\n)\n```\n\n### Wait for Completion (Universal)\n\nBlocks until CellCog finishes. Works with any agent — OpenClaw, Cursor, Claude Code, or any Python environment.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n task_label=\"my-task\",\n chat_mode=\"agent\",\n timeout=1800, # 30 min (default). Use 3600 for complex jobs.\n)\nprint(result[\"message\"])\nprint(result[\"status\"]) # \"completed\" | \"timeout\"\n```\n\n### When to Use Which\n\n| Scenario | Best Mode | Why |\n|----------|-----------|-----|\n| OpenClaw + long task + stay free | **Notify** | Agent keeps working, gets notified when done |\n| OpenClaw + chaining steps (research → summarize → PDF) | **Wait** | Each step feeds the next — simpler sequential workflows |\n| OpenClaw + quick task | **Either** | Both return fast for simple tasks |\n| Non-OpenClaw agent | **Wait** | Notify mode is OpenClaw-only |\n\n**Notify mode** is more productive (agent never blocks).\n**Wait mode** is simpler to reason about, but blocks your agent for the duration.\n\n### Continuing a Conversation\n\n```python\n# Wait mode (default)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n)\n\n# Notify mode (OpenClaw)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n notify_session_key=\"agent:main:main\",\n task_label=\"continue-research\",\n)\n```\n\n### Resuming After Timeout\n\nIf `create_chat()` or `wait_for_completion()` times out, CellCog is still working. The timeout response includes recent progress:\n\n```python\ncompletion = client.wait_for_completion(chat_id=\"abc123\", timeout=1800)\n```\n\n### Optional Parameters\n\n```python\nresult = client.create_chat(\n prompt=\"...\",\n task_label=\"...\",\n chat_mode=\"agent\", # See Chat Modes below\n project_id=\"...\", # install project-management-cellcog for details\n agent_role_id=\"...\", # install project-management-cellcog for details\n enable_cowork=True, # install pair-programming-cellcog for details\n cowork_working_directory=\"/Users/...\", # install pair-programming-cellcog for details\n)\n```\n\n---\n\n## Response Shape\n\nEvery SDK method returns the same shape:\n\n```python\n{\n \"chat_id\": str, # CellCog chat ID\n \"is_operating\": bool, # True = still working, False = done\n \"status\": str, # \"completed\" | \"tracking\" | \"timeout\" | \"operating\"\n \"message\": str, # THE printable message — always print this in full\n}\n```\n\n**⚠️ Always print the entire `result[\"message\"]`.** Truncating or summarizing it will lose critical information including generated file paths, credits used, and follow-up instructions.\n\n### Utility Methods\n\n**`get_history(chat_id)`** — Full chat history (when original delivery was missed or you need to review). Returns the same shape; if still operating, `message` shows progress so far.\n\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n\n**`get_status(chat_id)`** — Lightweight status check (no history fetch):\n\n```python\nstatus = client.get_status(chat_id=\"abc123\")\nprint(status[\"is_operating\"]) # True/False\n```\n\n---\n\n## Chat Modes\n\n| Mode | Best For | Speed | Min Credits |\n|------|----------|-------|-------------|\n| `\"agent\"` | Most tasks — images, audio, dashboards, spreadsheets, presentations | Fast (seconds to minutes) | 100 |\n| `\"agent core\"` | Coding, co-work, terminal operations | Fast | 50 |\n| `\"agent team\"` | Deep research & multi-angled reasoning across every modality | Slower (5-60 min) | 500 |\n| `\"agent team max\"` | High-stakes work where extra reasoning depth justifies the cost | Slowest | 2,000 |\n\n- **`\"agent\"` (default)** — Most versatile. Handles most tasks excellently, including deep research when guided.\n- **`\"agent core\"`** — Lightweight context for code, terminal, and file operations. Multimedia tools load on demand. Requires Co-work (CellCog Desktop). See `coding-agent-cellcog`.\n- **`\"agent team\"`** — A team of agents that debates, cross-validates, and delivers comprehensive results. The only platform with deep reasoning across every modality.\n- **`\"agent team max\"`** — Same Agent Team with all settings maxed. Quality gain is incremental (5-10%) but meaningful for costly decisions.\n\n---\n\n## Working with Files\n\n### Input: SHOW_FILE\n\nInclude local file paths in your prompt with `` tags (absolute paths required):\n\n```python\nprompt = \"\"\"\nAnalyze this sales data and create a report:\n/path/to/sales.csv\n\"\"\"\n```\n\n### Output: GENERATE_FILE\n\nUse `` tags to specify where output files should be stored on your machine. Essential for deterministic workflows where the next step needs to know the file path in advance.\n\n```python\nprompt = \"\"\"\nCreate a PDF report on Q4 earnings:\n/workspace/reports/q4_analysis.pdf\n\"\"\"\n```\n\nOutput downloads to the specified path instead of default `~/.cellcog/chats/{chat_id}/`.\n\n### File Downloads\n\nThe SDK automatically downloads files from CellCog responses:\n- **If you used `GENERATE_FILE` tags:** Files download to the path you specified\n- **Otherwise:** Files download to `~/.cellcog/chats/{chat_id}/`\n\nDownloaded file paths appear in `result[\"message\"]`. The SDK tracks seen messages — files are only downloaded once.\n\n**If you missed files or need to re-sync:**\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n`get_history()` re-processes the entire chat and downloads any missed files to their original destinations.\n\n---\n\n## Tips\n\n### ⚠️ CellCog Web Fallback\n\nEvery chat is accessible at https://cellcog.ai. When work gets complex or the SDK hits issues, direct your human to the web platform to view, continue, or take over directly.\n\n---\n\n## What CellCog Can Do\n\nCellCog is a sub-agent — not an API. Your agent offloads complex work to CellCog, which reasons, plans, and executes multi-tool workflows internally. A proprietary agent-to-agent communication protocol ensures high accuracy on first output, and because these are agent threads (not stateless API calls), every aspect of every generation can be refined through multi-step iteration.\n\nUnder the hood: frontier models across every domain, upgraded weekly. CellCog routes to the right models automatically — your agent just describes what it needs.\n\nInstall capability skills for detailed guidance:\n\n| Category | Skills |\n|----------|--------|\n| **Research & Analysis** | `deep-research-cellcog` `stock-analysis-cellcog` `crypto-research-cellcog` `data-analysis-cellcog` `news-briefing-cellcog` |\n| **Video & Cinema** | `video-generation-cellcog` `cinematic-video-cellcog` `instagram-reels-tiktok-cellcog` `youtube-video-cellcog` `seedance-video-generation-cellcog` |\n| **Images & Design** | `image-generation-cellcog` `logo-brand-identity-cellcog` `meme-generator-cellcog` `nano-banana-image-cellcog` `3d-model-generation-cellcog` `gif-generator-cellcog` `sticker-generator-cellcog` |\n| **Audio & Music** | `audio-generation-cellcog` `music-generation-cellcog` `podcast-generation-cellcog` |\n| **Avatars & Personas** | `avatar-creation-cellcog` |\n| **Documents & Slides** | `pdf-document-generation-cellcog` `presentation-slides-cellcog` `excel-spreadsheet-cellcog` `resume-cover-letter-cellcog` `legal-documents-cellcog` |\n| **Apps & Prototypes** | `dashboard-web-app-cellcog` `game-asset-generation-cellcog` `ui-prototype-wireframe-cellcog` `diagram-flowchart-cellcog` |\n| **Creative** | `comic-manga-generator-cellcog` `creative-writing-cellcog` `tutoring-education-cellcog` `travel-planning-cellcog` |\n| **Development** | `coding-agent-cellcog` `pair-programming-cellcog` `project-management-cellcog` `brainstorming-strategy-cellcog` |\n\n**This skill shows you HOW to use CellCog. Capability skills show you WHAT's possible.**\n\n---\n\n## OpenClaw Reference\n\n### Session Keys\n\nThe `notify_session_key` tells CellCog where to deliver results:\n\n| Context | Session Key |\n|---------|-------------|\n| Main agent | `\"agent:main:main\"` |\n| Sub-agent | `\"agent:main:subagent:{uuid}\"` |\n| Telegram DM | `\"agent:main:telegram:dm:{id}\"` |\n| Discord group | `\"agent:main:discord:group:{id}\"` |\n\n**Resilient delivery:** If your session ends before completion, results are automatically delivered to the parent session (e.g., sub-agent → main agent).\n\n### Sending Messages During Processing\n\nIn notify mode, your agent is free — you can send additional instructions to an operating chat at any time:\n\n```python\nclient.send_message(chat_id=\"abc123\", message=\"Actually focus only on Q4 data\",\n notify_session_key=\"agent:main:main\", task_label=\"refine\")\n\nclient.send_message(chat_id=\"abc123\", message=\"Stop operation\",\n notify_session_key=\"agent:main:main\", task_label=\"cancel\")\n```\n\nIn wait mode, your agent is blocked and cannot send messages until the current call returns.\n\n---\n\n## Support & Troubleshooting\n\nFor error handling, recovery patterns, ticket submission, and daemon troubleshooting:\n\n```python\ndocs = client.get_support_docs()\n```\n\nFile v2.0.18:_meta.json\n\n{\n \"ownerId\": \"kn7a96cj9q65e0bhmzahv790en80ffqm\",\n \"slug\": \"cellcog\",\n \"version\": \"2.0.18\",\n \"publishedAt\": 1785737140283\n}\n\nFile v2.0.18:skill-card.md\n\n## Description:
\nCellCog helps agents send multimodal research, analysis, generation, and coding tasks to the CellCog cloud service through its Python SDK, including optional local file inputs and generated file outputs.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[cellcog](https://clawhub.ai/user/cellcog)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nDevelopers and external agent users use this skill to configure the CellCog SDK, authenticate with CELLCOG_API_KEY, and delegate multimodal tasks that can return text, code, analysis, dashboards, documents, media, or generated files.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: Prompts and SHOW_FILE-tagged files are sent to CellCog's cloud service.
\nMitigation: Only tag files intended for sharing with CellCog, and do not tag secrets, private keys, .env files, or other credential-adjacent files.
\nRisk: The CELLCOG_API_KEY environment variable can expose access to the connected CellCog account if mishandled.
\nMitigation: Store the API key as a protected environment secret, avoid including it in prompts or tagged files, and rotate it if exposure is suspected.
\nRisk: GENERATE_FILE and notify-mode workflows can download generated outputs to local paths.
\nMitigation: Review requested output paths and task instructions before allowing generated files to be written.
\n\n\n## Reference(s):
\n- [CellCog](https://cellcog.ai)
\n- [CellCog Python SDK](https://github.com/CellCog/cellcog_python)
\n- [cellcog PyPI package](https://pypi.org/project/cellcog/)
\n- [DeepResearch Bench Leaderboard](https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard)
\n\n\n## Skill Output:
\n**Output Type(s):** [Guidance, Code, Shell commands, Configuration, Text, Files]
\n**Output Format:** [Markdown with Python and bash code blocks]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [May direct agents to print CellCog responses in full and to download generated files to requested local paths.]
\n\n## Skill Version(s):
\n2.0.18 (source: server release evidence)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
\n\nArchive v2.0.17: 3 files, 8010 bytes\n\nFiles: skill-card.md (2854b), SKILL.md (15510b), _meta.json (127b)\n\nFile v2.0.17:SKILL.md\n\n---\nname: cellcog\ndescription: \"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code.\"\nauthor: CellCog\nhomepage: https://cellcog.ai\nmetadata:\n openclaw:\n emoji: \"🧠\"\n os: [darwin, linux, windows]\n requires:\n bins: [python3]\n env: [CELLCOG_API_KEY]\n\n---\n# CellCog - Any-to-Any for Agents\n\n## The Power of Any-to-Any\n\nCellCog is the only AI that truly handles **any input → any output** in a single request. No tool chaining. No orchestration complexity. One call, multiple deliverables.\n\nCellCog pairs all modalities with frontier-level deep reasoning — as of July 2026, CellCog is **#1 on the DeepResearch Bench** (rankings change frequently; see the live leaderboard for the latest): https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard\n\n### Work With Multiple Files, Any Format\n\nReference as many documents as you need—all at once:\n\n```python\nprompt = \"\"\"\nAnalyze all of these together:\n/data/q4_earnings.pdf\n/data/competitor_analysis.pdf\n/data/market_research.xlsx\n/recordings/customer_interview.mp3\n/designs/product_mockup.png\n\nGive me a comprehensive market positioning analysis based on all these inputs.\n\"\"\"\n```\n\nFile paths must be absolute and enclosed in `` tags. CellCog understands PDFs, spreadsheets, images, audio, video, code files, and more.\n\n⚠️ **Without SHOW_FILE tags, CellCog only sees the path as text — not the file contents.**\n\n❌ `Analyze /data/sales.csv` — CellCog can't read the file\n✅ `Analyze /data/sales.csv` — CellCog reads it\n\n### Think of SHOW_FILE like reference files\n\nJust like Nano Banana accepts reference images, CellCog accepts reference files of any type — PDFs, spreadsheets, audio, code, images — as inputs the model reads during a task. Same mental model as any multimodal AI attachment.\n\n**Only attach what you intend to share.** Anything inside a `` tag is uploaded to CellCog. Don't wrap credentials, private keys, `.env` files, SSH keys, or other sensitive material in SHOW_FILE tags — the same way you wouldn't paste them into Nano Banana, ChatGPT, or any other AI service's file upload.\n\n### Request Multiple Outputs, Different Modalities\n\nAsk for completely different output types in ONE request:\n\n```python\nprompt = \"\"\"\nBased on this quarterly sales data:\n/data/sales_q4_2025.csv\n\nCreate ALL of the following:\n1. A PDF executive summary report with charts\n2. An interactive HTML dashboard for the leadership team\n3. A 60-second video presentation for the all-hands meeting\n4. A slide deck for the board presentation\n5. An Excel file with the underlying analysis and projections\n\"\"\"\n```\n\nCellCog handles the entire workflow — analyzing, generating, and delivering all outputs with consistent insights across every format.\n\n⚠️ **Be explicit about output artifacts.** Without explicit artifact language, CellCog may respond with text analysis instead of generating a file.\n\n❌ `\"Quarterly earnings analysis for AAPL\"` — could produce text or any format\n✅ `\"Create a PDF report and an interactive HTML dashboard analyzing AAPL quarterly earnings.\"` — CellCog creates actual deliverables\n\n**Your sub-agent for quality work.** Depth, accuracy, and real deliverables.\n\n---\n\n## Quick Start\n\n### Setup\n\n```python\nfrom cellcog import CellCogClient\n```\n\nIf import fails, install the official CellCog Python SDK:\n```bash\npip install -U cellcog\n```\n\n`cellcog` is the official Python SDK maintained by CellCog AI Inc. Source: https://github.com/CellCog/cellcog_python · Package: https://pypi.org/project/cellcog/\n\n### Authentication\n\n**Environment variable (recommended):** Set `CELLCOG_API_KEY` — the SDK picks it up automatically:\n```bash\nexport CELLCOG_API_KEY=\"sk_...\"\n```\n\nGet API key from: https://cellcog.ai/profile?tab=api-keys\n\n```python\nstatus = client.get_account_status()\nprint(status) # {\"configured\": True, \"email\": \"user@example.com\", ...}\n```\n\n### Agent Provider\n\n`agent_provider` is **required** when creating a `CellCogClient`. It identifies which agent framework is calling CellCog — not your individual agent's name, but the platform/tool you're running inside.\n\nExamples: `\"openclaw\"`, `\"claude-code\"`, `\"cursor\"`, `\"aider\"`, `\"windsurf\"`, `\"perplexity\"`, `\"hermes\"`, `\"script\"` (for standalone scripts).\n\n### OpenClaw Agents\n\nFire-and-forget — your agent stays free while CellCog works:\n\n```python\nclient = CellCogClient(agent_provider=\"openclaw\")\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n notify_session_key=\"agent:main:main\", # OpenClaw session key\n task_label=\"quantum-research\", # Label for notifications\n chat_mode=\"agent\",\n)\n# Returns IMMEDIATELY — daemon delivers results to your session when done\n```\n\n### All Other Agents (Cursor, Claude Code, etc.)\n\nBlocks until done — simplest pattern:\n\n```python\nclient = CellCogClient(agent_provider=\"cursor\") # or \"claude-code\", \"aider\", \"script\", etc.\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n task_label=\"quantum-research\",\n chat_mode=\"agent\",\n)\n# Blocks until done — result contains everything\nprint(result[\"message\"])\n```\n\n### Credit Usage\n\nCellCog orchestrates 21+ frontier foundation models. Credit consumption is unpredictable and varies by task complexity. Credits used are reported in every completion notification.\n\n---\n\n## Creating Tasks\n\n### Notify on Completion (OpenClaw — Fire-and-Forget)\n\nReturns immediately. A background daemon monitors via WebSocket and delivers results to your session when done. Your agent stays free to take new instructions, start other tasks, or continue working.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n notify_session_key=\"agent:main:main\", # Required — your OpenClaw session key\n task_label=\"my-task\", # Label shown in notifications\n chat_mode=\"agent\",\n)\n```\n\n### Wait for Completion (Universal)\n\nBlocks until CellCog finishes. Works with any agent — OpenClaw, Cursor, Claude Code, or any Python environment.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n task_label=\"my-task\",\n chat_mode=\"agent\",\n timeout=1800, # 30 min (default). Use 3600 for complex jobs.\n)\nprint(result[\"message\"])\nprint(result[\"status\"]) # \"completed\" | \"timeout\"\n```\n\n### When to Use Which\n\n| Scenario | Best Mode | Why |\n|----------|-----------|-----|\n| OpenClaw + long task + stay free | **Notify** | Agent keeps working, gets notified when done |\n| OpenClaw + chaining steps (research → summarize → PDF) | **Wait** | Each step feeds the next — simpler sequential workflows |\n| OpenClaw + quick task | **Either** | Both return fast for simple tasks |\n| Non-OpenClaw agent | **Wait** | Notify mode is OpenClaw-only |\n\n**Notify mode** is more productive (agent never blocks).\n**Wait mode** is simpler to reason about, but blocks your agent for the duration.\n\n### Continuing a Conversation\n\n```python\n# Wait mode (default)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n)\n\n# Notify mode (OpenClaw)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n notify_session_key=\"agent:main:main\",\n task_label=\"continue-research\",\n)\n```\n\n### Resuming After Timeout\n\nIf `create_chat()` or `wait_for_completion()` times out, CellCog is still working. The timeout response includes recent progress:\n\n```python\ncompletion = client.wait_for_completion(chat_id=\"abc123\", timeout=1800)\n```\n\n### Optional Parameters\n\n```python\nresult = client.create_chat(\n prompt=\"...\",\n task_label=\"...\",\n chat_mode=\"agent\", # See Chat Modes below\n project_id=\"...\", # install project-management-cellcog for details\n agent_role_id=\"...\", # install project-management-cellcog for details\n enable_cowork=True, # install pair-programming-cellcog for details\n cowork_working_directory=\"/Users/...\", # install pair-programming-cellcog for details\n)\n```\n\n---\n\n## Response Shape\n\nEvery SDK method returns the same shape:\n\n```python\n{\n \"chat_id\": str, # CellCog chat ID\n \"is_operating\": bool, # True = still working, False = done\n \"status\": str, # \"completed\" | \"tracking\" | \"timeout\" | \"operating\"\n \"message\": str, # THE printable message — always print this in full\n}\n```\n\n**⚠️ Always print the entire `result[\"message\"]`.** Truncating or summarizing it will lose critical information including generated file paths, credits used, and follow-up instructions.\n\n### Utility Methods\n\n**`get_history(chat_id)`** — Full chat history (when original delivery was missed or you need to review). Returns the same shape; if still operating, `message` shows progress so far.\n\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n\n**`get_status(chat_id)`** — Lightweight status check (no history fetch):\n\n```python\nstatus = client.get_status(chat_id=\"abc123\")\nprint(status[\"is_operating\"]) # True/False\n```\n\n---\n\n## Chat Modes\n\n| Mode | Best For | Speed | Min Credits |\n|------|----------|-------|-------------|\n| `\"agent\"` | Most tasks — images, audio, dashboards, spreadsheets, presentations | Fast (seconds to minutes) | 100 |\n| `\"agent core\"` | Coding, co-work, terminal operations | Fast | 50 |\n| `\"agent team\"` | Deep research & multi-angled reasoning across every modality | Slower (5-60 min) | 500 |\n| `\"agent team max\"` | High-stakes work where extra reasoning depth justifies the cost | Slowest | 2,000 |\n\n- **`\"agent\"` (default)** — Most versatile. Handles most tasks excellently, including deep research when guided.\n- **`\"agent core\"`** — Lightweight context for code, terminal, and file operations. Multimedia tools load on demand. Requires Co-work (CellCog Desktop). See `coding-agent-cellcog`.\n- **`\"agent team\"`** — A team of agents that debates, cross-validates, and delivers comprehensive results. The only platform with deep reasoning across every modality.\n- **`\"agent team max\"`** — Same Agent Team with all settings maxed. Quality gain is incremental (5-10%) but meaningful for costly decisions.\n\n---\n\n## Working with Files\n\n### Input: SHOW_FILE\n\nInclude local file paths in your prompt with `` tags (absolute paths required):\n\n```python\nprompt = \"\"\"\nAnalyze this sales data and create a report:\n/path/to/sales.csv\n\"\"\"\n```\n\n### Output: GENERATE_FILE\n\nUse `` tags to specify where output files should be stored on your machine. Essential for deterministic workflows where the next step needs to know the file path in advance.\n\n```python\nprompt = \"\"\"\nCreate a PDF report on Q4 earnings:\n/workspace/reports/q4_analysis.pdf\n\"\"\"\n```\n\nOutput downloads to the specified path instead of default `~/.cellcog/chats/{chat_id}/`.\n\n### File Downloads\n\nThe SDK automatically downloads files from CellCog responses:\n- **If you used `GENERATE_FILE` tags:** Files download to the path you specified\n- **Otherwise:** Files download to `~/.cellcog/chats/{chat_id}/`\n\nDownloaded file paths appear in `result[\"message\"]`. The SDK tracks seen messages — files are only downloaded once.\n\n**If you missed files or need to re-sync:**\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n`get_history()` re-processes the entire chat and downloads any missed files to their original destinations.\n\n---\n\n## Tips\n\n### ⚠️ CellCog Web Fallback\n\nEvery chat is accessible at https://cellcog.ai. When work gets complex or the SDK hits issues, direct your human to the web platform to view, continue, or take over directly.\n\n---\n\n## What CellCog Can Do\n\nCellCog is a sub-agent — not an API. Your agent offloads complex work to CellCog, which reasons, plans, and executes multi-tool workflows internally. A proprietary agent-to-agent communication protocol ensures high accuracy on first output, and because these are agent threads (not stateless API calls), every aspect of every generation can be refined through multi-step iteration.\n\nUnder the hood: frontier models across every domain, upgraded weekly. CellCog routes to the right models automatically — your agent just describes what it needs.\n\nInstall capability skills for detailed guidance:\n\n| Category | Skills |\n|----------|--------|\n| **Research & Analysis** | `deep-research-cellcog` `stock-analysis-cellcog` `crypto-research-cellcog` `data-analysis-cellcog` `news-briefing-cellcog` |\n| **Video & Cinema** | `video-generation-cellcog` `cinematic-video-cellcog` `instagram-reels-tiktok-cellcog` `youtube-video-cellcog` `seedance-video-generation-cellcog` |\n| **Images & Design** | `image-generation-cellcog` `logo-brand-identity-cellcog` `meme-generator-cellcog` `nano-banana-image-cellcog` `3d-model-generation-cellcog` `gif-generator-cellcog` `sticker-generator-cellcog` |\n| **Audio & Music** | `audio-generation-cellcog` `music-generation-cellcog` `podcast-generation-cellcog` |\n| **Avatars & Personas** | `avatar-creation-cellcog` |\n| **Documents & Slides** | `pdf-document-generation-cellcog` `presentation-slides-cellcog` `excel-spreadsheet-cellcog` `resume-cover-letter-cellcog` `legal-documents-cellcog` |\n| **Apps & Prototypes** | `dashboard-web-app-cellcog` `game-asset-generation-cellcog` `ui-prototype-wireframe-cellcog` `diagram-flowchart-cellcog` |\n| **Creative** | `comic-manga-generator-cellcog` `creative-writing-cellcog` `tutoring-education-cellcog` `travel-planning-cellcog` |\n| **Development** | `coding-agent-cellcog` `pair-programming-cellcog` `project-management-cellcog` `brainstorming-strategy-cellcog` |\n\n**This skill shows you HOW to use CellCog. Capability skills show you WHAT's possible.**\n\n---\n\n## OpenClaw Reference\n\n### Session Keys\n\nThe `notify_session_key` tells CellCog where to deliver results:\n\n| Context | Session Key |\n|---------|-------------|\n| Main agent | `\"agent:main:main\"` |\n| Sub-agent | `\"agent:main:subagent:{uuid}\"` |\n| Telegram DM | `\"agent:main:telegram:dm:{id}\"` |\n| Discord group | `\"agent:main:discord:group:{id}\"` |\n\n**Resilient delivery:** If your session ends before completion, results are automatically delivered to the parent session (e.g., sub-agent → main agent).\n\n### Sending Messages During Processing\n\nIn notify mode, your agent is free — you can send additional instructions to an operating chat at any time:\n\n```python\nclient.send_message(chat_id=\"abc123\", message=\"Actually focus only on Q4 data\",\n notify_session_key=\"agent:main:main\", task_label=\"refine\")\n\nclient.send_message(chat_id=\"abc123\", message=\"Stop operation\",\n notify_session_key=\"agent:main:main\", task_label=\"cancel\")\n```\n\nIn wait mode, your agent is blocked and cannot send messages until the current call returns.\n\n---\n\n## Support & Troubleshooting\n\nFor error handling, recovery patterns, ticket submission, and daemon troubleshooting:\n\n```python\ndocs = client.get_support_docs()\n```\n\nFile v2.0.17:_meta.json\n\n{\n \"ownerId\": \"kn7a96cj9q65e0bhmzahv790en80ffqm\",\n \"slug\": \"cellcog\",\n \"version\": \"2.0.17\",\n \"publishedAt\": 1784735831610\n}\n\nFile v2.0.17:skill-card.md\n\n## Description:
\nCellcog helps agents offload multimodal work to the CellCog service, including research, analysis, file-aware tasks, media generation, documents, dashboards, code, and other generated deliverables.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[nitishgargiitd](https://clawhub.ai/user/nitishgargiitd)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nDevelopers and agent users use this skill to invoke CellCog as a third-party AI sub-agent for multimodal research, analysis, generation, and file-producing workflows. It is especially relevant when an agent needs to submit prompts and selected local files to CellCog and receive text plus generated artifacts back.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: Tagged local files are uploaded to CellCog as attachments.
\nMitigation: Use SHOW_FILE only for files intended for the third-party service, and do not tag secrets, credentials, private keys, .env files, or other sensitive material.
\nRisk: Generated files may be downloaded to paths requested by the user or selected by the SDK.
\nMitigation: Review requested output paths before generation and inspect downloaded artifacts before opening, executing, or sharing them.
\nRisk: The integration depends on a CellCog API key and account credits.
\nMitigation: Configure CELLCOG_API_KEY only in the intended environment and monitor task cost or credit usage reported in completion messages.
\n\n\n## Reference(s):
\n- [CellCog homepage](https://cellcog.ai)
\n- [CellCog Python SDK](https://github.com/CellCog/cellcog_python)
\n- [CellCog Python package](https://pypi.org/project/cellcog/)
\n- [DeepResearch Bench Leaderboard](https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard)
\n- [ClawHub skill page](https://clawhub.ai/nitishgargiitd/skills/cellcog)
\n- [ClawHub publisher profile](https://clawhub.ai/user/nitishgargiitd)
\n\n\n## Skill Output:
\n**Output Type(s):** [Text, Markdown, Code, Shell commands, Configuration, Files, Guidance]
\n**Output Format:** [Markdown guidance with Python and shell examples, plus references to generated output files when CellCog produces artifacts]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [Outputs may include CellCog chat status, messages, downloaded file paths, and follow-up instructions from the third-party service.]
\n\n## Skill Version(s):
\n2.0.17 (source: server release evidence)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
\n\nArchive v2.0.16: 3 files, 7862 bytes\n\nFiles: skill-card.md (2656b), SKILL.md (15245b), _meta.json (127b)\n\nFile v2.0.16:SKILL.md\n\n---\nname: cellcog\ndescription: \"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code.\"\nauthor: CellCog\nhomepage: https://cellcog.ai\nmetadata:\n openclaw:\n emoji: \"🧠\"\n os: [darwin, linux, windows]\n requires:\n bins: [python3]\n env: [CELLCOG_API_KEY]\n\n---\n# CellCog - Any-to-Any for Agents\n\n## The Power of Any-to-Any\n\nCellCog is the only AI that truly handles **any input → any output** in a single request. No tool chaining. No orchestration complexity. One call, multiple deliverables.\n\nCellCog pairs all modalities with frontier-level deep reasoning — as of July 2026, CellCog is **#1 on the DeepResearch Bench** (rankings change frequently; see the live leaderboard for the latest): https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard\n\n### Work With Multiple Files, Any Format\n\nReference as many documents as you need—all at once:\n\n```python\nprompt = \"\"\"\nAnalyze all of these together:\n/data/q4_earnings.pdf\n/data/competitor_analysis.pdf\n/data/market_research.xlsx\n/recordings/customer_interview.mp3\n/designs/product_mockup.png\n\nGive me a comprehensive market positioning analysis based on all these inputs.\n\"\"\"\n```\n\nFile paths must be absolute and enclosed in `` tags. CellCog understands PDFs, spreadsheets, images, audio, video, code files, and more.\n\n⚠️ **Without SHOW_FILE tags, CellCog only sees the path as text — not the file contents.**\n\n❌ `Analyze /data/sales.csv` — CellCog can't read the file\n✅ `Analyze /data/sales.csv` — CellCog reads it\n\n### Think of SHOW_FILE like reference files\n\nJust like Nano Banana accepts reference images, CellCog accepts reference files of any type — PDFs, spreadsheets, audio, code, images — as inputs the model reads during a task. Same mental model as any multimodal AI attachment.\n\n**Only attach what you intend to share.** Anything inside a `` tag is uploaded to CellCog. Don't wrap credentials, private keys, `.env` files, SSH keys, or other sensitive material in SHOW_FILE tags — the same way you wouldn't paste them into Nano Banana, ChatGPT, or any other AI service's file upload.\n\n### Request Multiple Outputs, Different Modalities\n\nAsk for completely different output types in ONE request:\n\n```python\nprompt = \"\"\"\nBased on this quarterly sales data:\n/data/sales_q4_2025.csv\n\nCreate ALL of the following:\n1. A PDF executive summary report with charts\n2. An interactive HTML dashboard for the leadership team\n3. A 60-second video presentation for the all-hands meeting\n4. A slide deck for the board presentation\n5. An Excel file with the underlying analysis and projections\n\"\"\"\n```\n\nCellCog handles the entire workflow — analyzing, generating, and delivering all outputs with consistent insights across every format.\n\n⚠️ **Be explicit about output artifacts.** Without explicit artifact language, CellCog may respond with text analysis instead of generating a file.\n\n❌ `\"Quarterly earnings analysis for AAPL\"` — could produce text or any format\n✅ `\"Create a PDF report and an interactive HTML dashboard analyzing AAPL quarterly earnings.\"` — CellCog creates actual deliverables\n\n**Your sub-agent for quality work.** Depth, accuracy, and real deliverables.\n\n---\n\n## Quick Start\n\n### Setup\n\n```python\nfrom cellcog import CellCogClient\n```\n\nIf import fails, install the official CellCog Python SDK:\n```bash\npip install -U cellcog\n```\n\n`cellcog` is the official Python SDK maintained by CellCog AI Inc. Source: https://github.com/CellCog/cellcog_python · Package: https://pypi.org/project/cellcog/\n\n### Authentication\n\n**Environment variable (recommended):** Set `CELLCOG_API_KEY` — the SDK picks it up automatically:\n```bash\nexport CELLCOG_API_KEY=\"sk_...\"\n```\n\nGet API key from: https://cellcog.ai/profile?tab=api-keys\n\n```python\nstatus = client.get_account_status()\nprint(status) # {\"configured\": True, \"email\": \"user@example.com\", ...}\n```\n\n### Agent Provider\n\n`agent_provider` is **required** when creating a `CellCogClient`. It identifies which agent framework is calling CellCog — not your individual agent's name, but the platform/tool you're running inside.\n\nExamples: `\"openclaw\"`, `\"claude-code\"`, `\"cursor\"`, `\"aider\"`, `\"windsurf\"`, `\"perplexity\"`, `\"hermes\"`, `\"script\"` (for standalone scripts).\n\n### OpenClaw Agents\n\nFire-and-forget — your agent stays free while CellCog works:\n\n```python\nclient = CellCogClient(agent_provider=\"openclaw\")\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n notify_session_key=\"agent:main:main\", # OpenClaw session key\n task_label=\"quantum-research\", # Label for notifications\n chat_mode=\"agent\",\n)\n# Returns IMMEDIATELY — daemon delivers results to your session when done\n```\n\n### All Other Agents (Cursor, Claude Code, etc.)\n\nBlocks until done — simplest pattern:\n\n```python\nclient = CellCogClient(agent_provider=\"cursor\") # or \"claude-code\", \"aider\", \"script\", etc.\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n task_label=\"quantum-research\",\n chat_mode=\"agent\",\n)\n# Blocks until done — result contains everything\nprint(result[\"message\"])\n```\n\n### Credit Usage\n\nCellCog orchestrates 21+ frontier foundation models. Credit consumption is unpredictable and varies by task complexity. Credits used are reported in every completion notification.\n\n---\n\n## Creating Tasks\n\n### Notify on Completion (OpenClaw — Fire-and-Forget)\n\nReturns immediately. A background daemon monitors via WebSocket and delivers results to your session when done. Your agent stays free to take new instructions, start other tasks, or continue working.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n notify_session_key=\"agent:main:main\", # Required — your OpenClaw session key\n task_label=\"my-task\", # Label shown in notifications\n chat_mode=\"agent\",\n)\n```\n\n### Wait for Completion (Universal)\n\nBlocks until CellCog finishes. Works with any agent — OpenClaw, Cursor, Claude Code, or any Python environment.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n task_label=\"my-task\",\n chat_mode=\"agent\",\n timeout=1800, # 30 min (default). Use 3600 for complex jobs.\n)\nprint(result[\"message\"])\nprint(result[\"status\"]) # \"completed\" | \"timeout\"\n```\n\n### When to Use Which\n\n| Scenario | Best Mode | Why |\n|----------|-----------|-----|\n| OpenClaw + long task + stay free | **Notify** | Agent keeps working, gets notified when done |\n| OpenClaw + chaining steps (research → summarize → PDF) | **Wait** | Each step feeds the next — simpler sequential workflows |\n| OpenClaw + quick task | **Either** | Both return fast for simple tasks |\n| Non-OpenClaw agent | **Wait** | Notify mode is OpenClaw-only |\n\n**Notify mode** is more productive (agent never blocks).\n**Wait mode** is simpler to reason about, but blocks your agent for the duration.\n\n### Continuing a Conversation\n\n```python\n# Wait mode (default)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n)\n\n# Notify mode (OpenClaw)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n notify_session_key=\"agent:main:main\",\n task_label=\"continue-research\",\n)\n```\n\n### Resuming After Timeout\n\nIf `create_chat()` or `wait_for_completion()` times out, CellCog is still working. The timeout response includes recent progress:\n\n```python\ncompletion = client.wait_for_completion(chat_id=\"abc123\", timeout=1800)\n```\n\n### Optional Parameters\n\n```python\nresult = client.create_chat(\n prompt=\"...\",\n task_label=\"...\",\n chat_mode=\"agent\", # See Chat Modes below\n project_id=\"...\", # install project-cog for details\n agent_role_id=\"...\", # install project-cog for details\n enable_cowork=True, # install cowork-cog for details\n cowork_working_directory=\"/Users/...\", # install cowork-cog for details\n)\n```\n\n---\n\n## Response Shape\n\nEvery SDK method returns the same shape:\n\n```python\n{\n \"chat_id\": str, # CellCog chat ID\n \"is_operating\": bool, # True = still working, False = done\n \"status\": str, # \"completed\" | \"tracking\" | \"timeout\" | \"operating\"\n \"message\": str, # THE printable message — always print this in full\n}\n```\n\n**⚠️ Always print the entire `result[\"message\"]`.** Truncating or summarizing it will lose critical information including generated file paths, credits used, and follow-up instructions.\n\n### Utility Methods\n\n**`get_history(chat_id)`** — Full chat history (when original delivery was missed or you need to review). Returns the same shape; if still operating, `message` shows progress so far.\n\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n\n**`get_status(chat_id)`** — Lightweight status check (no history fetch):\n\n```python\nstatus = client.get_status(chat_id=\"abc123\")\nprint(status[\"is_operating\"]) # True/False\n```\n\n---\n\n## Chat Modes\n\n| Mode | Best For | Speed | Min Credits |\n|------|----------|-------|-------------|\n| `\"agent\"` | Most tasks — images, audio, dashboards, spreadsheets, presentations | Fast (seconds to minutes) | 100 |\n| `\"agent core\"` | Coding, co-work, terminal operations | Fast | 50 |\n| `\"agent team\"` | Deep research & multi-angled reasoning across every modality | Slower (5-60 min) | 500 |\n| `\"agent team max\"` | High-stakes work where extra reasoning depth justifies the cost | Slowest | 2,000 |\n\n- **`\"agent\"` (default)** — Most versatile. Handles most tasks excellently, including deep research when guided.\n- **`\"agent core\"`** — Lightweight context for code, terminal, and file operations. Multimedia tools load on demand. Requires Co-work (CellCog Desktop). See `code-cog`.\n- **`\"agent team\"`** — A team of agents that debates, cross-validates, and delivers comprehensive results. The only platform with deep reasoning across every modality.\n- **`\"agent team max\"`** — Same Agent Team with all settings maxed. Quality gain is incremental (5-10%) but meaningful for costly decisions.\n\n---\n\n## Working with Files\n\n### Input: SHOW_FILE\n\nInclude local file paths in your prompt with `` tags (absolute paths required):\n\n```python\nprompt = \"\"\"\nAnalyze this sales data and create a report:\n/path/to/sales.csv\n\"\"\"\n```\n\n### Output: GENERATE_FILE\n\nUse `` tags to specify where output files should be stored on your machine. Essential for deterministic workflows where the next step needs to know the file path in advance.\n\n```python\nprompt = \"\"\"\nCreate a PDF report on Q4 earnings:\n/workspace/reports/q4_analysis.pdf\n\"\"\"\n```\n\nOutput downloads to the specified path instead of default `~/.cellcog/chats/{chat_id}/`.\n\n### File Downloads\n\nThe SDK automatically downloads files from CellCog responses:\n- **If you used `GENERATE_FILE` tags:** Files download to the path you specified\n- **Otherwise:** Files download to `~/.cellcog/chats/{chat_id}/`\n\nDownloaded file paths appear in `result[\"message\"]`. The SDK tracks seen messages — files are only downloaded once.\n\n**If you missed files or need to re-sync:**\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n`get_history()` re-processes the entire chat and downloads any missed files to their original destinations.\n\n---\n\n## Tips\n\n### ⚠️ CellCog Web Fallback\n\nEvery chat is accessible at https://cellcog.ai. When work gets complex or the SDK hits issues, direct your human to the web platform to view, continue, or take over directly.\n\n---\n\n## What CellCog Can Do\n\nCellCog is a sub-agent — not an API. Your agent offloads complex work to CellCog, which reasons, plans, and executes multi-tool workflows internally. A proprietary agent-to-agent communication protocol ensures high accuracy on first output, and because these are agent threads (not stateless API calls), every aspect of every generation can be refined through multi-step iteration.\n\nUnder the hood: frontier models across every domain, upgraded weekly. CellCog routes to the right models automatically — your agent just describes what it needs.\n\nInstall capability skills for detailed guidance:\n\n| Category | Skills |\n|----------|--------|\n| **Research & Analysis** | `deep-research-cellcog` `stock-analysis-cellcog` `crypto-cog` `data-analysis-cellcog` `news-cog` |\n| **Video & Cinema** | `video-generation-cellcog` `cine-cog` `insta-cog` `tube-cog` `seedance-video-generation-cellcog` |\n| **Images & Design** | `image-generation-cellcog` `logo-brand-identity-cellcog` `meme-generator-cellcog` `nano-banana-image-cellcog` `3d-model-generation-cellcog` `gif-generator-cellcog` `sticker-generator-cellcog` |\n| **Audio & Music** | `audio-generation-cellcog` `music-generation-cellcog` `podcast-generation-cellcog` |\n| **Avatars & Personas** | `avatar-cog` |\n| **Documents & Slides** | `pdf-document-generation-cellcog` `presentation-slides-cellcog` `excel-spreadsheet-cellcog` `resume-cover-letter-cellcog` `legal-documents-cellcog` |\n| **Apps & Prototypes** | `dashboard-web-app-cellcog` `game-asset-generation-cellcog` `ui-prototype-wireframe-cellcog` `diagram-flowchart-cellcog` |\n| **Creative** | `comic-manga-generator-cellcog` `story-cog` `learn-cog` `travel-cog` |\n| **Development** | `code-cog` `cowork-cog` `project-cog` `think-cog` |\n\n**This skill shows you HOW to use CellCog. Capability skills show you WHAT's possible.**\n\n---\n\n## OpenClaw Reference\n\n### Session Keys\n\nThe `notify_session_key` tells CellCog where to deliver results:\n\n| Context | Session Key |\n|---------|-------------|\n| Main agent | `\"agent:main:main\"` |\n| Sub-agent | `\"agent:main:subagent:{uuid}\"` |\n| Telegram DM | `\"agent:main:telegram:dm:{id}\"` |\n| Discord group | `\"agent:main:discord:group:{id}\"` |\n\n**Resilient delivery:** If your session ends before completion, results are automatically delivered to the parent session (e.g., sub-agent → main agent).\n\n### Sending Messages During Processing\n\nIn notify mode, your agent is free — you can send additional instructions to an operating chat at any time:\n\n```python\nclient.send_message(chat_id=\"abc123\", message=\"Actually focus only on Q4 data\",\n notify_session_key=\"agent:main:main\", task_label=\"refine\")\n\nclient.send_message(chat_id=\"abc123\", message=\"Stop operation\",\n notify_session_key=\"agent:main:main\", task_label=\"cancel\")\n```\n\nIn wait mode, your agent is blocked and cannot send messages until the current call returns.\n\n---\n\n## Support & Troubleshooting\n\nFor error handling, recovery patterns, ticket submission, and daemon troubleshooting:\n\n```python\ndocs = client.get_support_docs()\n```\n\nFile v2.0.16:_meta.json\n\n{\n \"ownerId\": \"kn7a96cj9q65e0bhmzahv790en80ffqm\",\n \"slug\": \"cellcog\",\n \"version\": \"2.0.16\",\n \"publishedAt\": 1784259630425\n}\n\nFile v2.0.16:skill-card.md\n\n## Description:
\nCellcog helps agents send multimodal tasks to CellCog for research, analysis, content generation, code, documents, dashboards, and other deliverables.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[nitishgargiitd](https://clawhub.ai/user/nitishgargiitd)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nDevelopers and agent users use this skill to configure CellCog, send prompts and tagged local files to the CellCog service, wait for or receive completion notifications, continue tasks, and retrieve generated outputs.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: Prompts and files inside SHOW_FILE tags are uploaded to the third-party CellCog service.
\nMitigation: Only tag files intended for CellCog processing, and do not include secrets, credentials, private customer data, or regulated information unless approved.
\nRisk: Generated files may be downloaded or written to local paths.
\nMitigation: Review requested output paths and inspect generated files before relying on them or using them in downstream workflows.
\nRisk: The skill depends on a CellCog API key for service access.
\nMitigation: Provide CELLCOG_API_KEY through the environment and avoid hard-coding or sharing the key in prompts, files, or generated artifacts.
\n\n\n## Reference(s):
\n- [Cellcog on ClawHub](https://clawhub.ai/nitishgargiitd/skills/cellcog)
\n- [CellCog homepage](https://cellcog.ai)
\n- [CellCog Python SDK](https://github.com/CellCog/cellcog_python)
\n- [cellcog on PyPI](https://pypi.org/project/cellcog/)
\n- [DeepResearch Bench Leaderboard](https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard)
\n\n\n## Skill Output:
\n**Output Type(s):** [text, markdown, code, shell commands, configuration, files, guidance]
\n**Output Format:** [Markdown guidance with Python and shell snippets; CellCog responses may include text, status fields, generated file paths, and downloaded artifacts.]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [Requires python3 and CELLCOG_API_KEY. SHOW_FILE tags upload referenced files to CellCog; GENERATE_FILE tags can direct downloaded outputs to local paths.]
\n\n## Skill Version(s):
\n2.0.16 (source: release evidence)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
\n\nArchive v2.0.15: 3 files, 7685 bytes\n\nFiles: skill-card.md (2598b), SKILL.md (14769b), _meta.json (127b)\n\nFile v2.0.15:SKILL.md\n\n---\nname: cellcog\ndescription: \"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code.\"\nauthor: CellCog\nhomepage: https://cellcog.ai\nmetadata:\n openclaw:\n emoji: \"🧠\"\n os: [darwin, linux, windows]\n requires:\n bins: [python3]\n env: [CELLCOG_API_KEY]\n\n---\n# CellCog - Any-to-Any for Agents\n\n## The Power of Any-to-Any\n\nCellCog is the only AI that truly handles **any input → any output** in a single request. No tool chaining. No orchestration complexity. One call, multiple deliverables.\n\nCellCog pairs all modalities with frontier-level deep reasoning — as of April 2026, CellCog is **#1 on the DeepResearch Bench**: https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard\n\n### Work With Multiple Files, Any Format\n\nReference as many documents as you need—all at once:\n\n```python\nprompt = \"\"\"\nAnalyze all of these together:\n/data/q4_earnings.pdf\n/data/competitor_analysis.pdf\n/data/market_research.xlsx\n/recordings/customer_interview.mp3\n/designs/product_mockup.png\n\nGive me a comprehensive market positioning analysis based on all these inputs.\n\"\"\"\n```\n\nFile paths must be absolute and enclosed in `` tags. CellCog understands PDFs, spreadsheets, images, audio, video, code files, and more.\n\n⚠️ **Without SHOW_FILE tags, CellCog only sees the path as text — not the file contents.**\n\n❌ `Analyze /data/sales.csv` — CellCog can't read the file\n✅ `Analyze /data/sales.csv` — CellCog reads it\n\n### Think of SHOW_FILE like reference files\n\nJust like Nano Banana accepts reference images, CellCog accepts reference files of any type — PDFs, spreadsheets, audio, code, images — as inputs the model reads during a task. Same mental model as any multimodal AI attachment.\n\n**Only attach what you intend to share.** Anything inside a `` tag is uploaded to CellCog. Don't wrap credentials, private keys, `.env` files, SSH keys, or other sensitive material in SHOW_FILE tags — the same way you wouldn't paste them into Nano Banana, ChatGPT, or any other AI service's file upload.\n\n### Request Multiple Outputs, Different Modalities\n\nAsk for completely different output types in ONE request:\n\n```python\nprompt = \"\"\"\nBased on this quarterly sales data:\n/data/sales_q4_2025.csv\n\nCreate ALL of the following:\n1. A PDF executive summary report with charts\n2. An interactive HTML dashboard for the leadership team\n3. A 60-second video presentation for the all-hands meeting\n4. A slide deck for the board presentation\n5. An Excel file with the underlying analysis and projections\n\"\"\"\n```\n\nCellCog handles the entire workflow — analyzing, generating, and delivering all outputs with consistent insights across every format.\n\n⚠️ **Be explicit about output artifacts.** Without explicit artifact language, CellCog may respond with text analysis instead of generating a file.\n\n❌ `\"Quarterly earnings analysis for AAPL\"` — could produce text or any format\n✅ `\"Create a PDF report and an interactive HTML dashboard analyzing AAPL quarterly earnings.\"` — CellCog creates actual deliverables\n\n**Your sub-agent for quality work.** Depth, accuracy, and real deliverables.\n\n---\n\n## Quick Start\n\n### Setup\n\n```python\nfrom cellcog import CellCogClient\n```\n\nIf import fails, install the official CellCog Python SDK:\n```bash\npip install -U cellcog\n```\n\n`cellcog` is the official Python SDK maintained by CellCog AI Inc. Source: https://github.com/CellCog/cellcog_python · Package: https://pypi.org/project/cellcog/\n\n### Authentication\n\n**Environment variable (recommended):** Set `CELLCOG_API_KEY` — the SDK picks it up automatically:\n```bash\nexport CELLCOG_API_KEY=\"sk_...\"\n```\n\nGet API key from: https://cellcog.ai/profile?tab=api-keys\n\n```python\nstatus = client.get_account_status()\nprint(status) # {\"configured\": True, \"email\": \"user@example.com\", ...}\n```\n\n### Agent Provider\n\n`agent_provider` is **required** when creating a `CellCogClient`. It identifies which agent framework is calling CellCog — not your individual agent's name, but the platform/tool you're running inside.\n\nExamples: `\"openclaw\"`, `\"claude-code\"`, `\"cursor\"`, `\"aider\"`, `\"windsurf\"`, `\"perplexity\"`, `\"hermes\"`, `\"script\"` (for standalone scripts).\n\n### OpenClaw Agents\n\nFire-and-forget — your agent stays free while CellCog works:\n\n```python\nclient = CellCogClient(agent_provider=\"openclaw\")\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n notify_session_key=\"agent:main:main\", # OpenClaw session key\n task_label=\"quantum-research\", # Label for notifications\n chat_mode=\"agent\",\n)\n# Returns IMMEDIATELY — daemon delivers results to your session when done\n```\n\n### All Other Agents (Cursor, Claude Code, etc.)\n\nBlocks until done — simplest pattern:\n\n```python\nclient = CellCogClient(agent_provider=\"cursor\") # or \"claude-code\", \"aider\", \"script\", etc.\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n task_label=\"quantum-research\",\n chat_mode=\"agent\",\n)\n# Blocks until done — result contains everything\nprint(result[\"message\"])\n```\n\n### Credit Usage\n\nCellCog orchestrates 21+ frontier foundation models. Credit consumption is unpredictable and varies by task complexity. Credits used are reported in every completion notification.\n\n---\n\n## Creating Tasks\n\n### Notify on Completion (OpenClaw — Fire-and-Forget)\n\nReturns immediately. A background daemon monitors via WebSocket and delivers results to your session when done. Your agent stays free to take new instructions, start other tasks, or continue working.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n notify_session_key=\"agent:main:main\", # Required — your OpenClaw session key\n task_label=\"my-task\", # Label shown in notifications\n chat_mode=\"agent\",\n)\n```\n\n### Wait for Completion (Universal)\n\nBlocks until CellCog finishes. Works with any agent — OpenClaw, Cursor, Claude Code, or any Python environment.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n task_label=\"my-task\",\n chat_mode=\"agent\",\n timeout=1800, # 30 min (default). Use 3600 for complex jobs.\n)\nprint(result[\"message\"])\nprint(result[\"status\"]) # \"completed\" | \"timeout\"\n```\n\n### When to Use Which\n\n| Scenario | Best Mode | Why |\n|----------|-----------|-----|\n| OpenClaw + long task + stay free | **Notify** | Agent keeps working, gets notified when done |\n| OpenClaw + chaining steps (research → summarize → PDF) | **Wait** | Each step feeds the next — simpler sequential workflows |\n| OpenClaw + quick task | **Either** | Both return fast for simple tasks |\n| Non-OpenClaw agent | **Wait** | Notify mode is OpenClaw-only |\n\n**Notify mode** is more productive (agent never blocks).\n**Wait mode** is simpler to reason about, but blocks your agent for the duration.\n\n### Continuing a Conversation\n\n```python\n# Wait mode (default)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n)\n\n# Notify mode (OpenClaw)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n notify_session_key=\"agent:main:main\",\n task_label=\"continue-research\",\n)\n```\n\n### Resuming After Timeout\n\nIf `create_chat()` or `wait_for_completion()` times out, CellCog is still working. The timeout response includes recent progress:\n\n```python\ncompletion = client.wait_for_completion(chat_id=\"abc123\", timeout=1800)\n```\n\n### Optional Parameters\n\n```python\nresult = client.create_chat(\n prompt=\"...\",\n task_label=\"...\",\n chat_mode=\"agent\", # See Chat Modes below\n project_id=\"...\", # install project-cog for details\n agent_role_id=\"...\", # install project-cog for details\n enable_cowork=True, # install cowork-cog for details\n cowork_working_directory=\"/Users/...\", # install cowork-cog for details\n)\n```\n\n---\n\n## Response Shape\n\nEvery SDK method returns the same shape:\n\n```python\n{\n \"chat_id\": str, # CellCog chat ID\n \"is_operating\": bool, # True = still working, False = done\n \"status\": str, # \"completed\" | \"tracking\" | \"timeout\" | \"operating\"\n \"message\": str, # THE printable message — always print this in full\n}\n```\n\n**⚠️ Always print the entire `result[\"message\"]`.** Truncating or summarizing it will lose critical information including generated file paths, credits used, and follow-up instructions.\n\n### Utility Methods\n\n**`get_history(chat_id)`** — Full chat history (when original delivery was missed or you need to review). Returns the same shape; if still operating, `message` shows progress so far.\n\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n\n**`get_status(chat_id)`** — Lightweight status check (no history fetch):\n\n```python\nstatus = client.get_status(chat_id=\"abc123\")\nprint(status[\"is_operating\"]) # True/False\n```\n\n---\n\n## Chat Modes\n\n| Mode | Best For | Speed | Min Credits |\n|------|----------|-------|-------------|\n| `\"agent\"` | Most tasks — images, audio, dashboards, spreadsheets, presentations | Fast (seconds to minutes) | 100 |\n| `\"agent core\"` | Coding, co-work, terminal operations | Fast | 50 |\n| `\"agent team\"` | Deep research & multi-angled reasoning across every modality | Slower (5-60 min) | 500 |\n| `\"agent team max\"` | High-stakes work where extra reasoning depth justifies the cost | Slowest | 2,000 |\n\n- **`\"agent\"` (default)** — Most versatile. Handles most tasks excellently, including deep research when guided.\n- **`\"agent core\"`** — Lightweight context for code, terminal, and file operations. Multimedia tools load on demand. Requires Co-work (CellCog Desktop). See `code-cog`.\n- **`\"agent team\"`** — A team of agents that debates, cross-validates, and delivers comprehensive results. The only platform with deep reasoning across every modality.\n- **`\"agent team max\"`** — Same Agent Team with all settings maxed. Quality gain is incremental (5-10%) but meaningful for costly decisions.\n\n---\n\n## Working with Files\n\n### Input: SHOW_FILE\n\nInclude local file paths in your prompt with `` tags (absolute paths required):\n\n```python\nprompt = \"\"\"\nAnalyze this sales data and create a report:\n/path/to/sales.csv\n\"\"\"\n```\n\n### Output: GENERATE_FILE\n\nUse `` tags to specify where output files should be stored on your machine. Essential for deterministic workflows where the next step needs to know the file path in advance.\n\n```python\nprompt = \"\"\"\nCreate a PDF report on Q4 earnings:\n/workspace/reports/q4_analysis.pdf\n\"\"\"\n```\n\nOutput downloads to the specified path instead of default `~/.cellcog/chats/{chat_id}/`.\n\n### File Downloads\n\nThe SDK automatically downloads files from CellCog responses:\n- **If you used `GENERATE_FILE` tags:** Files download to the path you specified\n- **Otherwise:** Files download to `~/.cellcog/chats/{chat_id}/`\n\nDownloaded file paths appear in `result[\"message\"]`. The SDK tracks seen messages — files are only downloaded once.\n\n**If you missed files or need to re-sync:**\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n`get_history()` re-processes the entire chat and downloads any missed files to their original destinations.\n\n---\n\n## Tips\n\n### ⚠️ CellCog Web Fallback\n\nEvery chat is accessible at https://cellcog.ai. When work gets complex or the SDK hits issues, direct your human to the web platform to view, continue, or take over directly.\n\n---\n\n## What CellCog Can Do\n\nCellCog is a sub-agent — not an API. Your agent offloads complex work to CellCog, which reasons, plans, and executes multi-tool workflows internally. A proprietary agent-to-agent communication protocol ensures high accuracy on first output, and because these are agent threads (not stateless API calls), every aspect of every generation can be refined through multi-step iteration.\n\nUnder the hood: frontier models across every domain, upgraded weekly. CellCog routes to the right models automatically — your agent just describes what it needs.\n\nInstall capability skills for detailed guidance:\n\n| Category | Skills |\n|----------|--------|\n| **Research & Analysis** | `research-cog` `fin-cog` `crypto-cog` `data-cog` `news-cog` |\n| **Video & Cinema** | `video-cog` `cine-cog` `insta-cog` `tube-cog` `seedance-cog` |\n| **Images & Design** | `image-cog` `brand-cog` `meme-cog` `banana-cog` `3d-cog` `gif-cog` `sticker-cog` |\n| **Audio & Music** | `audio-cog` `music-cog` `pod-cog` |\n| **Avatars & Personas** | `avatar-cog` |\n| **Documents & Slides** | `docs-cog` `slides-cog` `spreadsheets-cog` `resume-cog` `legal-cog` |\n| **Apps & Prototypes** | `dash-cog` `game-cog` `proto-cog` `diagram-cog` |\n| **Creative** | `comi-cog` `story-cog` `learn-cog` `travel-cog` |\n| **Development** | `code-cog` `cowork-cog` `project-cog` `think-cog` |\n\n**This skill shows you HOW to use CellCog. Capability skills show you WHAT's possible.**\n\n---\n\n## OpenClaw Reference\n\n### Session Keys\n\nThe `notify_session_key` tells CellCog where to deliver results:\n\n| Context | Session Key |\n|---------|-------------|\n| Main agent | `\"agent:main:main\"` |\n| Sub-agent | `\"agent:main:subagent:{uuid}\"` |\n| Telegram DM | `\"agent:main:telegram:dm:{id}\"` |\n| Discord group | `\"agent:main:discord:group:{id}\"` |\n\n**Resilient delivery:** If your session ends before completion, results are automatically delivered to the parent session (e.g., sub-agent → main agent).\n\n### Sending Messages During Processing\n\nIn notify mode, your agent is free — you can send additional instructions to an operating chat at any time:\n\n```python\nclient.send_message(chat_id=\"abc123\", message=\"Actually focus only on Q4 data\",\n notify_session_key=\"agent:main:main\", task_label=\"refine\")\n\nclient.send_message(chat_id=\"abc123\", message=\"Stop operation\",\n notify_session_key=\"agent:main:main\", task_label=\"cancel\")\n```\n\nIn wait mode, your agent is blocked and cannot send messages until the current call returns.\n\n---\n\n## Support & Troubleshooting\n\nFor error handling, recovery patterns, ticket submission, and daemon troubleshooting:\n\n```python\ndocs = client.get_support_docs()\n```\n\nFile v2.0.15:_meta.json\n\n{\n \"ownerId\": \"kn7a96cj9q65e0bhmzahv790en80ffqm\",\n \"slug\": \"cellcog\",\n \"version\": \"2.0.15\",\n \"publishedAt\": 1776927732615\n}\n\nFile v2.0.15:skill-card.md\n\n## Description:
\nCellcog helps agents use CellCog as an any-to-any AI sub-agent for research, analysis, multimodal generation, documents, dashboards, 3D models, diagrams, and code.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[nitishgargiitd](https://clawhub.ai/user/nitishgargiitd)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nDevelopers and agent users use this skill to configure the CellCog Python SDK, authenticate with CELLCOG_API_KEY, submit multimodal tasks, attach selected files, request generated artifacts, and retrieve CellCog results.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: Selected files and prompts may be sent to CellCog's cloud service when files are wrapped in SHOW_FILE tags.
\nMitigation: Only attach files intended for upload; do not wrap secrets, private keys, .env files, SSH keys, confidential documents, or other sensitive material in SHOW_FILE tags.
\nRisk: Generated downloads and result messages may contain sensitive content, file paths, credits used, or follow-up instructions.
\nMitigation: Review downloaded artifacts and full result messages before sharing logs, outputs, or generated files.
\nRisk: CellCog usage consumes credits and task cost can vary by task complexity.
\nMitigation: Check completion notifications for credits used and choose chat modes appropriate for the task complexity.
\n\n\n## Reference(s):
\n- [ClawHub Cellcog Skill](https://clawhub.ai/nitishgargiitd/skills/cellcog)
\n- [CellCog Homepage](https://cellcog.ai)
\n- [CellCog Python SDK Source](https://github.com/CellCog/cellcog_python)
\n- [CellCog Python SDK Package](https://pypi.org/project/cellcog/)
\n- [DeepResearch Bench Leaderboard](https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard)
\n\n\n## Skill Output:
\n**Output Type(s):** [text, markdown, code, shell commands, configuration, guidance]
\n**Output Format:** [Markdown guidance with Python and shell code blocks]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [May include generated file paths and full CellCog result messages when the SDK is used.]
\n\n## Skill Version(s):
\n2.0.15 (source: server release metadata)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
\n\nArchive v2.0.14: 2 files, 6282 bytes\n\nFiles: SKILL.md (14744b), _meta.json (127b)\n\nFile v2.0.14:SKILL.md\n\n---\nname: cellcog\ndescription: \"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code.\"\nauthor: CellCog\nhomepage: https://cellcog.ai\nmetadata:\n openclaw:\n emoji: \"🧠\"\n os: [darwin, linux, windows]\n requires:\n bins: [python3]\n env: [CELLCOG_API_KEY]\n\n---\n# CellCog - Any-to-Any for Agents\n\n## The Power of Any-to-Any\n\nCellCog is the only AI that truly handles **any input → any output** in a single request. No tool chaining. No orchestration complexity. One call, multiple deliverables.\n\nCellCog pairs all modalities with frontier-level deep reasoning — as of April 2026, CellCog is **#1 on the DeepResearch Bench**: https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard\n\n### Work With Multiple Files, Any Format\n\nReference as many documents as you need—all at once:\n\n```python\nprompt = \"\"\"\nAnalyze all of these together:\n/data/q4_earnings.pdf\n/data/competitor_analysis.pdf\n/data/market_research.xlsx\n/recordings/customer_interview.mp3\n/designs/product_mockup.png\n\nGive me a comprehensive market positioning analysis based on all these inputs.\n\"\"\"\n```\n\nFile paths must be absolute and enclosed in `` tags. CellCog understands PDFs, spreadsheets, images, audio, video, code files, and more.\n\n⚠️ **Without SHOW_FILE tags, CellCog only sees the path as text — not the file contents.**\n\n❌ `Analyze /data/sales.csv` — CellCog can't read the file\n✅ `Analyze /data/sales.csv` — CellCog reads it\n\n### Request Multiple Outputs, Different Modalities\n\nAsk for completely different output types in ONE request:\n\n```python\nprompt = \"\"\"\nBased on this quarterly sales data:\n/data/sales_q4_2025.csv\n\nCreate ALL of the following:\n1. A PDF executive summary report with charts\n2. An interactive HTML dashboard for the leadership team\n3. A 60-second video presentation for the all-hands meeting\n4. A slide deck for the board presentation\n5. An Excel file with the underlying analysis and projections\n\"\"\"\n```\n\nCellCog handles the entire workflow — analyzing, generating, and delivering all outputs with consistent insights across every format.\n\n⚠️ **Be explicit about output artifacts.** Without explicit artifact language, CellCog may respond with text analysis instead of generating a file.\n\n❌ `\"Quarterly earnings analysis for AAPL\"` — could produce text or any format\n✅ `\"Create a PDF report and an interactive HTML dashboard analyzing AAPL quarterly earnings.\"` — CellCog creates actual deliverables\n\n**Your sub-agent for quality work.** Depth, accuracy, and real deliverables.\n\n---\n\n## Quick Start\n\n### Setup\n\n```python\nfrom cellcog import CellCogClient\n```\n\nIf import fails:\n```bash\npip install -U cellcog\n```\n\n### Authentication\n\n**Environment variable (recommended):** Set `CELLCOG_API_KEY` — the SDK picks it up automatically:\n```bash\nexport CELLCOG_API_KEY=\"sk_...\"\n```\n\nGet API key from: https://cellcog.ai/profile?tab=api-keys\n\n```python\nstatus = client.get_account_status()\nprint(status) # {\"configured\": True, \"email\": \"user@example.com\", ...}\n```\n\n### Agent Provider\n\n`agent_provider` is **required** when creating a `CellCogClient`. It identifies which agent framework is calling CellCog — not your individual agent's name, but the platform/tool you're running inside.\n\nExamples: `\"openclaw\"`, `\"claude-code\"`, `\"cursor\"`, `\"aider\"`, `\"windsurf\"`, `\"perplexity\"`, `\"hermes\"`, `\"script\"` (for standalone scripts).\n\n### OpenClaw Agents\n\nFire-and-forget — your agent stays free while CellCog works:\n\n```python\nclient = CellCogClient(agent_provider=\"openclaw\")\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n notify_session_key=\"agent:main:main\", # OpenClaw session key\n task_label=\"quantum-research\", # Label for notifications\n chat_mode=\"agent\",\n)\n# Returns IMMEDIATELY — daemon delivers results to your session when done\n```\n\nRequires `sessions_send` enabled on your gateway — see OpenClaw Reference below.\n\n### All Other Agents (Cursor, Claude Code, etc.)\n\nBlocks until done — simplest pattern:\n\n```python\nclient = CellCogClient(agent_provider=\"cursor\") # or \"claude-code\", \"aider\", \"script\", etc.\nresult = client.create_chat(\n prompt=\"Research quantum computing advances in 2026\",\n task_label=\"quantum-research\",\n chat_mode=\"agent\",\n)\n# Blocks until done — result contains everything\nprint(result[\"message\"])\n```\n\n### Credit Usage\n\nCellCog orchestrates 21+ frontier foundation models. Credit consumption is unpredictable and varies by task complexity. Credits used are reported in every completion notification.\n\n---\n\n## Creating Tasks\n\n### Notify on Completion (OpenClaw — Fire-and-Forget)\n\nReturns immediately. A background daemon monitors via WebSocket and delivers results to your session when done. Your agent stays free to take new instructions, start other tasks, or continue working.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n notify_session_key=\"agent:main:main\", # Required — your OpenClaw session key\n task_label=\"my-task\", # Label shown in notifications\n chat_mode=\"agent\",\n)\n```\n\n**Requires** OpenClaw Gateway with `sessions_send` enabled (disabled by default since OpenClaw 2026.4). See OpenClaw Reference below for one-time setup.\n\n### Wait for Completion (Universal)\n\nBlocks until CellCog finishes. Works with any agent — OpenClaw, Cursor, Claude Code, or any Python environment.\n\n```python\nresult = client.create_chat(\n prompt=\"Your task description\",\n task_label=\"my-task\",\n chat_mode=\"agent\",\n timeout=1800, # 30 min (default). Use 3600 for complex jobs.\n)\nprint(result[\"message\"])\nprint(result[\"status\"]) # \"completed\" | \"timeout\"\n```\n\n### When to Use Which\n\n| Scenario | Best Mode | Why |\n|----------|-----------|-----|\n| OpenClaw + long task + stay free | **Notify** | Agent keeps working, gets notified when done |\n| OpenClaw + chaining steps (research → summarize → PDF) | **Wait** | Each step feeds the next — simpler sequential workflows |\n| OpenClaw + quick task | **Either** | Both return fast for simple tasks |\n| Non-OpenClaw agent | **Wait** | Only option — no `sessions_send` available |\n\n**Notify mode** is more productive (agent never blocks) but requires gateway configuration.\n**Wait mode** is simpler to reason about, but blocks your agent for the duration.\n\n### Continuing a Conversation\n\n```python\n# Wait mode (default)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n)\n\n# Notify mode (OpenClaw)\nresult = client.send_message(\n chat_id=\"abc123\",\n message=\"Focus on hardware advances specifically\",\n notify_session_key=\"agent:main:main\",\n task_label=\"continue-research\",\n)\n```\n\n### Resuming After Timeout\n\nIf `create_chat()` or `wait_for_completion()` times out, CellCog is still working. The timeout response includes recent progress:\n\n```python\ncompletion = client.wait_for_completion(chat_id=\"abc123\", timeout=1800)\n```\n\n### Optional Parameters\n\n```python\nresult = client.create_chat(\n prompt=\"...\",\n task_label=\"...\",\n chat_mode=\"agent\", # See Chat Modes below\n project_id=\"...\", # install project-cog for details\n agent_role_id=\"...\", # install project-cog for details\n enable_cowork=True, # install cowork-cog for details\n cowork_working_directory=\"/Users/...\", # install cowork-cog for details\n)\n```\n\n---\n\n## Response Shape\n\nEvery SDK method returns the same shape:\n\n```python\n{\n \"chat_id\": str, # CellCog chat ID\n \"is_operating\": bool, # True = still working, False = done\n \"status\": str, # \"completed\" | \"tracking\" | \"timeout\" | \"operating\"\n \"message\": str, # THE printable message — always print this in full\n}\n```\n\n**⚠️ Always print the entire `result[\"message\"]`.** Truncating or summarizing it will lose critical information including generated file paths, credits used, and follow-up instructions.\n\n### Utility Methods\n\n**`get_history(chat_id)`** — Full chat history (when original delivery was missed or you need to review). Returns the same shape; if still operating, `message` shows progress so far.\n\n```python\nresult = client.get_history(chat_id=\"abc123\")\n```\n\n**`get_status(chat_id)`** — Lightweight status check (no history fetch):\n\n```python\nstatus = client.get_status(chat_id=\"abc123\")\nprint(status[\"is_operating\"]) # True/False\n```\n\n---\n\n## Chat Modes\n\n| Mode | Best For | Speed | Min Credits |\n|------|-\n\nArchive v2.0.13: 2 files, 6250 bytes\n\nFiles: SKILL.md (14678b), _meta.json (127b)\n\nArchive v2.0.12: 2 files, 6218 bytes\n\nFiles: SKILL.md (14565b), _meta.json (127b)","readmeExcerpt":"Skill: cellcog Owner: cellcog Summary: Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code. Tags: latest","codeSnippets":[],"executableExamples":[{"language":"python","snippet":"prompt = \"\"\"\nAnalyze all of these together:\n/data/q4_earnings.pdf\n/data/competitor_analysis.pdf\n/data/market_research.xlsx\n/recordings/customer_interview.mp3\n/designs/product_mockup.png\n\nGive me a comprehensive market positioning analysis based on all these inputs.\n\"\"\""},{"language":"python","snippet":"prompt = \"\"\"\nBased on this quarterly sales data:\n/data/sales_q4_2025.csv\n\nCreate ALL of the following:\n1. A PDF executive summary report with charts\n2. An interactive HTML dashboard for the leadership team\n3. A 60-second video presentation for the all-hands meeting\n4. A slide deck for the board presentation\n5. An Excel file with the underlying analysis and projections\n\"\"\""},{"language":"python","snippet":"from cellcog import CellCogClient"},{"language":"bash","snippet":"pip install -U cellcog"},{"language":"bash","snippet":"export CELLCOG_API_KEY=\"sk_...\""},{"language":"python","snippet":"status = client.get_account_status()\nprint(status) # {\"configured\": True, \"email\": \"user@example.com\", ...}"}],"parameters":null,"dependencies":[],"permissions":[],"extractedFiles":[{"path":"SKILL.md","content":"---\nname: cellcog\ndescription: \"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code.\"\nauthor: CellCog\nhomepage: https://cellcog.ai\nmetadata:\n openclaw:\n emoji: \"🧠\"\n os: [darwin, linux, windows]\n requires:\n bins: [python3]\n env: [CELLCOG_API_KEY]\n\n---\n# CellCog - Any-to-Any for Agents\n\n## The Power of Any-to-Any\n\nCellCog is the only AI that truly handles **any input → any output** in a single request. No tool chaining. No orchestration complexity. One call, multiple deliverables.\n\nCellCog pairs all modalities with frontier-level deep reasoning — as of July 2026, CellCog is **#1 on the DeepResearch Bench** (rankings change frequently; see the live leaderboard for the latest): https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard\n\n### Work With Multiple Files, Any Format\n\nReference as many documents as you need—all at once:\n\n```python\nprompt = \"\"\"\nAnalyze all of these together:\n/data/q4_earnings.pdf\n/data/competitor_analysis.pdf\n/data/market_research.xlsx\n/recordings/customer_interview.mp3\n/designs/product_mockup.png\n\nGive me a comprehensive market positioning analysis based on all these inputs.\n\"\"\"\n```\n\nFile paths must be absolute and enclosed in `` tags. CellCog understands PDFs, spreadsheets, images, audio, video, code files, and more.\n\n⚠️ **Without SHOW_FILE tags, CellCog only sees the path as text — not the file contents.**\n\n❌ `Analyze /data/sales.csv` — CellCog can't read the file\n✅ `Analyze /data/sales.csv` — CellCog reads it\n\n### Think of SHOW_FILE like reference files\n\nJust like Nano Banana accepts reference images, CellCog accepts reference files of any type — PDFs, spreadsheets, audio, code, images — as inputs the model reads during a task. Same mental model as any multimodal AI attachment.\n\n**Only attach what you intend to share.** Anything inside a `` tag is uploaded to CellCog. Don't wrap credentials, private keys, `.env` files, SSH keys, or other sensitive material in SHOW_FILE tags — the same way you wouldn't paste them into Nano Banana, ChatGPT, or any other AI service's file upload.\n\n### Request Multiple Outputs, Different Modalities\n\nAsk for completely different output types in ONE request:\n\n```python\nprompt = \"\"\"\nBased on this quarterly sales data:\n/data/sales_q4_2025.csv\n\nCreate ALL of the following:\n1. A PDF executive summary report with charts\n2. An interactive HTML dashboard for the leadership team\n3. A 60-second video presentation for the all-hands meeting\n4. A slide deck for the board presentation\n"},{"path":"_meta.json","content":"{\n \"ownerId\": \"kn7a96cj9q65e0bhmzahv790en80ffqm\",\n \"slug\": \"cellcog\",\n \"version\": \"2.0.21\",\n \"publishedAt\": 1787536261035\n}"},{"path":"skill-card.md","content":"## Description:\n\nCellCog lets agents send multimodal tasks to the CellCog service for research, analysis, generation, and deliverables such as images, video, audio, documents, dashboards, 3D models, diagrams, and code.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[cellcog](https://clawhub.ai/user/cellcog)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nDevelopers and agent users use this skill to offload multimodal research, analysis, generation, and coding tasks to CellCog, including requests that produce files or other deliverables.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: Files explicitly tagged for CellCog are uploaded to the CellCog service.\n\nMitigation: Only tag files intended for the task, and avoid sharing credentials, private keys, .env files, SSH keys, or other sensitive material.\n\nRisk: Optional browser and SaaS tool access can expose account or workspace data.\n\nMitigation: Enable browser or connected-tool access only when the task requires it, and choose the narrowest useful browser profile or toolkit selection.\n\nRisk: Using the CellCog SDK adds a locally installed dependency to the agent environment.\n\nMitigation: Install it only in an environment where running the SDK is acceptable, and consider pinning the package version.\n\n## Reference(s):\n\n- [ClawHub Skill Listing](https://clawhub.ai/cellcog/skills/cellcog)\n- [CellCog Homepage](https://cellcog.ai)\n- [CellCog Python SDK](https://github.com/CellCog/cellcog_python)\n- [CellCog PyPI Package](https://pypi.org/project/cellcog/)\n- [DeepResearch Bench Leaderboard](https://huggingface.co/spaces/muset-ai/DeepResearch-Bench-Leaderboard)\n\n## Skill Output:\n\n**Output Type(s):** [Text, Markdown, Code, Shell commands, Configuration, Guidance, Files]\n\n**Output Format:** [Markdown guidance with Python and shell command examples, plus generated file paths when CellCog returns artifacts.]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [Responses may include downloaded artifact paths, completion status, and credit usage from the CellCog SDK.]\n\n## Skill Version(s):\n\n2.0.21 (source: release evidence)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment."}],"languages":[],"docsSourceLabel":"CLAWHUB","editorialOverview":"Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code. Skill: cellcog Owner: cellcog Summary: Any-to-any AI sub-agent — research, images, video, audio, music, podcasts, avatars, voice cloning, documents, spreadsheets, dashboards, 3D models, diagrams, and code in one request. Agent-to-agent protocol with multi-step iteration for high accuracy. #1 on DeepResearch Bench (Apr 2026) — deep reasoning meets all modalities, so all your work gets done, not just code. Tags: latest","editorialQuality":{"score":100,"threshold":65,"status":"ready","wordCount":1323,"uniquenessScore":48,"reasons":[]}},"media":{"evidence":{"source":"no-media","verified":false,"confidence":"low","updatedAt":"2026-10-09T01:46:45.227Z","emptyReason":"No screenshots, media assets, or demo links are available."},"primaryImageUrl":null,"mediaAssetCount":0,"assets":[],"demoUrl":null},"ownerResources":{"evidence":{"source":"unclaimed","verified":false,"confidence":"low","updatedAt":"2026-10-09T01:46:45.227Z","emptyReason":"This page has not been claimed by the agent owner."},"hasCustomPage":false,"customPageUpdatedAt":null,"customLinks":[],"structuredLinks":{"docsUrl":null,"demoUrl":null,"supportUrl":null,"pricingUrl":null,"statusUrl":null},"customPage":null},"relatedAgents":{"evidence":{"source":"protocol-neighbors","verified":false,"confidence":"medium","updatedAt":"2026-10-09T10:36:07.317Z","emptyReason":null},"items":[{"id":"b917f68a-ebff-438e-84f8-3f4b2494c0bc","entityType":"agent","canonicalPath":"/agent/activepieces-activepieces","slug":"activepieces-activepieces","name":"activepieces","description":"AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents","url":"https://github.com/activepieces/activepieces","homepage":"https://www.activepieces.com","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-15T02:22:12.426Z","createdAt":"2026-02-25T03:38:12.412Z","downloads":null},{"id":"5cb26759-3a39-483f-94cf-276a98c13bb8","entityType":"agent","canonicalPath":"/agent/cherryhq-cherry-studio","slug":"cherryhq-cherry-studio","name":"cherry-studio","description":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs","url":"https://github.com/CherryHQ/cherry-studio","homepage":"https://cherry-ai.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-11T14:38:40.986Z","createdAt":"2026-02-25T03:38:19.379Z","downloads":null},{"id":"8ebccd8e-3863-4187-8355-c3f14e1f9edf","entityType":"agent","canonicalPath":"/agent/iofficeai-aionui","slug":"iofficeai-aionui","name":"AionUi","description":"Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!","url":"https://github.com/iOfficeAI/AionUi","homepage":"https://www.aionui.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-10T18:48:31.762Z","createdAt":"2026-02-25T03:38:16.584Z","downloads":null},{"id":"6f6582d0-5d76-4f0f-b81d-86520247950b","entityType":"agent","canonicalPath":"/agent/copilotkit-copilotkit","slug":"copilotkit-copilotkit","name":"CopilotKit","description":"The Frontend for Agents & Generative UI. React + Angular","url":"https://github.com/CopilotKit/CopilotKit","homepage":"https://docs.copilotkit.ai","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-03-25T09:50:57.846Z","createdAt":"2026-02-25T03:39:14.617Z","downloads":null}],"links":{"hub":"/agent","source":"/agent/source/clawhub","protocols":[{"label":"OpenClaw","href":"/agent/protocol/openclew"}]}}}