{"id":"84de3e5d-6b8b-4e37-9a40-50a01264e191","entityType":"agent","slug":"clawhub-jernejcicorbin-hub-claim-verifier","name":"Pre-Publish Fact Checker","canonicalUrl":"https://www.xpersona.co/agent/clawhub-jernejcicorbin-hub-claim-verifier","canonicalPath":"/agent/clawhub-jernejcicorbin-hub-claim-verifier","generatedAt":"2026-10-10T02:41:24.706Z","source":"CLAWHUB","claimStatus":"UNCLAIMED","verificationTier":"NONE","summary":{"evidence":{"source":"editorial-content","verified":true,"confidence":"high","updatedAt":"2026-10-09T02:01:37.520Z","emptyReason":null},"description":"Verify external factual claims in draft content before publishing or sending. Produces a structured claim verification report with evidence links. Skill: Pre-Publish Fact Checker Owner: jernejcicorbin-hub Summary: Verify external factual claims in draft content before publishing or sending. Produces a structured claim verification report with evidence links. Tags: latest:1.0.2 Version history: v1.0.2 | 2026-05-11T15:04:59.696Z | user No user-facing changes in this version. - Version bumped to 1.0.2, but SKILL.md and all files remain unchanged. v1.0.1 | 2026-05-","descriptionLabel":"Technical summary","evidenceSummary":"Capability contract not published. No trust telemetry is available yet. 13.5K downloads reported by the source. Last updated 10/9/2026.","installCommand":"clawhub skill install s1737s8ysxd4yb5mfz4hhg5v8h83tm05:claim-verifier","sourceUrl":"https://clawhub.ai/jernejcicorbin-hub/claim-verifier","homepage":"https://clawhub.ai/jernejcicorbin-hub/skills/claim-verifier","primaryLinks":[{"label":"View on ClawHub","url":"https://clawhub.ai/jernejcicorbin-hub/claim-verifier","kind":"source"},{"label":"Homepage","url":"https://clawhub.ai/jernejcicorbin-hub/skills/claim-verifier","kind":"homepage"}],"safetyScore":84,"overallRank":62,"popularityScore":83,"trustScore":null,"claimedByName":null,"isOwner":false,"seoDescription":"Verify external factual claims in draft content before publishing or sending. Produces a structured claim verification report with evidence links. Skill: Pre-Pu"},"coverage":{"evidence":{"source":"public-profile","verified":false,"confidence":"medium","updatedAt":"2026-10-09T02:01:37.520Z","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":"2026-10-09T02:01:37.520Z","emptyReason":null},"stars":null,"forks":null,"downloads":13484,"packageName":null,"latestVersion":"1.0.2","tractionLabel":"13.5K downloads"},"release":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-09T02:01:37.520Z","emptyReason":null},"lastUpdatedAt":"2026-10-09T02:01:37.520Z","lastCrawledAt":"2026-10-09T02:01:37.520Z","lastIndexedAt":null,"nextCrawlAt":"2026-10-10T02:01:37.520Z","lastVerifiedAt":null,"highlights":[{"version":"1.0.2","createdAt":"2026-05-11T15:04:59.696Z","changelog":"No user-facing changes in this version. - Version bumped to 1.0.2, but SKILL.md and all files remain unchanged.","fileCount":7,"zipByteSize":10236},{"version":"1.0.1","createdAt":"2026-05-11T15:03:26.882Z","changelog":"No user-visible changes in this version. - Version number updated to 1.0.1, but no code or documentation changes detected. - Functionality and usage remain the same as the previous release.","fileCount":6,"zipByteSize":8648},{"version":"1.0.0","createdAt":"2026-05-11T14:56:14.605Z","changelog":"Initial release of claim-verifier. - Fact-checks draft content for external factual claims before publishing or sending. - Produces a structured verification report with confidence scores, claim statuses, and evidence links. - Uses Prismfy as the primary search mechanism for claim verification. - Supports up to 25 claims per run; configurable strict mode and recency window. - Outputs concise chat verdicts plus optional machine-readable JSON reports. - Handles search fallback, evidence scoring, and clear failure codes for transparency and reliability.","fileCount":6,"zipByteSize":8494}]},"execution":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No published capability contract is available yet."},"installCommand":"clawhub skill install s1737s8ysxd4yb5mfz4hhg5v8h83tm05:claim-verifier","setupComplexity":"low","setupSteps":["Setup complexity is LOW. This package is likely designed for quick installation with minimal external side-effects.","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-jernejcicorbin-hub-claim-verifier/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-jernejcicorbin-hub-claim-verifier/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-jernejcicorbin-hub-claim-verifier/trust"},"curlExamples":["curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-jernejcicorbin-hub-claim-verifier/snapshot\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-jernejcicorbin-hub-claim-verifier/contract\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-jernejcicorbin-hub-claim-verifier/trust\""],"jsonRequestTemplate":{"query":"summarize this repo","constraints":{"maxLatencyMs":2000,"protocolPreference":["OPENCLEW"]}},"jsonResponseTemplate":{"ok":true,"result":{"summary":"...","confidence":0.9},"meta":{"source":"CLAWHUB","generatedAt":"2026-10-10T02:41:24.705Z"}},"retryPolicy":{"maxAttempts":3,"backoffMs":[500,1500,3500],"retryableConditions":["HTTP_429","HTTP_503","NETWORK_TIMEOUT"]}},"endpoints":{"dossierUrl":"https://www.xpersona.co/api/v1/agents/clawhub-jernejcicorbin-hub-claim-verifier/dossier","snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-jernejcicorbin-hub-claim-verifier/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-jernejcicorbin-hub-claim-verifier/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-jernejcicorbin-hub-claim-verifier/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":"2026-10-09T02:01:37.520Z","emptyReason":null},"readme":"Skill: Pre-Publish Fact Checker\n\nOwner: jernejcicorbin-hub\n\nSummary: Verify external factual claims in draft content before publishing or sending. Produces a structured claim verification report with evidence links.\n\nTags: latest:1.0.2\n\nVersion history:\n\nv1.0.2 | 2026-05-11T15:04:59.696Z | user\n\nNo user-facing changes in this version.  \n- Version bumped to 1.0.2, but SKILL.md and all files remain unchanged.\n\nv1.0.1 | 2026-05-11T15:03:26.882Z | user\n\nNo user-visible changes in this version.\n\n- Version number updated to 1.0.1, but no code or documentation changes detected.\n- Functionality and usage remain the same as the previous release.\n\nv1.0.0 | 2026-05-11T14:56:14.605Z | user\n\nInitial release of claim-verifier.\n\n- Fact-checks draft content for external factual claims before publishing or sending.\n- Produces a structured verification report with confidence scores, claim statuses, and evidence links.\n- Uses Prismfy as the primary search mechanism for claim verification.\n- Supports up to 25 claims per run; configurable strict mode and recency window.\n- Outputs concise chat verdicts plus optional machine-readable JSON reports.\n- Handles search fallback, evidence scoring, and clear failure codes for transparency and reliability.\n\nArchive index:\n\nArchive v1.0.2: 7 files, 10236 bytes\n\nFiles: _meta.json (133b), claim-verify-batch.sh (4357b), claim-verify.sh (2633b), hooks/claim-verifier/handler.js (1364b), hooks/claim-verifier/HOOK.md (531b), skill-card.md (2744b), SKILL.md (9321b)\n\nFile v1.0.2:SKILL.md\n\n---\nname: claim-verifier\ndescription: >\n  Verify external factual claims in draft content before publishing or sending.\n  Produces a structured claim verification report with evidence links.\nversion: 1.0.0\nmetadata:\n  openclaw:\n    requires:\n      env:\n        - PRISMFY_API_KEY\n      bins:\n        - curl\n        - jq\n    primaryEnv: PRISMFY_API_KEY\n    emoji: \"✅\"\n    homepage: https://prismfy.io\n---\n\n# Claim Verifier\n\nUse this skill to fact-check drafts before you publish or send them.\n\nWhy it helps:\n- catches false, stale, or conflicting external claims,\n- gives evidence links for each checked claim,\n- produces a publish-ready verification report you can trust.\n\nBest for:\n- blog posts,\n- product pages,\n- social posts,\n- outbound messages with external facts.\n\nNeed a key?\n- Get free search access and your API key at: https://prismfy.io\n\n## Setup\n\n1. Install the skill:\n```bash\nopenclaw skills install claim-verifier\n```\n\n2. Add your Prismfy API key:\n```bash\nexport PRISMFY_API_KEY=\"ss_live_your_key_here\"\n```\n\nTo keep it after restart:\n```bash\necho 'export PRISMFY_API_KEY=\"ss_live_your_key_here\"' >> ~/.bashrc\nsource ~/.bashrc\n```\n\n3. Verify API access:\n```bash\ncd ~/.openclaw/workspace/skills/claim-verifier\nbash claim-verify.sh --quota\n```\n\n4. Quick smoke test:\n```bash\ncd ~/.openclaw/workspace/skills/claim-verifier\nbash claim-verify.sh --claim \"OpenAI released GPT-5 in March 2026\"\n```\n\nEnable hook (recommended):\n```bash\n# Run from this skill directory:\n# ~/.openclaw/workspace/skills/claim-verifier\n\n# Copy OpenClaw hook into hooks registry\ncp -r hooks/claim-verifier ~/.openclaw/hooks/\n\n# Confirm files are present\nfind ~/.openclaw/hooks/claim-verifier -maxdepth 1 -type f | sort\n\n# Enable hook\nopenclaw hooks enable claim-verifier\n\n# Verify\nopenclaw hooks list\n```\n\nQuick example:\n\nDraft text:\n```text\nCompany X launched Feature Y in March 2026 and price starts at $49.\nThe company says this is the fastest tool in its category.\n```\n\nWhat Claim Verifier checks:\n1. `Company X launched Feature Y in March 2026`\n2. `Price starts at $49`\n3. `Fastest tool in its category`\n\nExample output (summary):\n```text\n1) verified     confidence: 0.91   evidence: 3 links\n2) conflicting  confidence: 0.58   evidence: 2 links (pricing mismatch)\n3) weak         confidence: 0.42   evidence: marketing claim, no independent benchmark\n```\n\nResult:\n- you immediately see what can stay,\n- what must be edited,\n- and which claims need stronger sources before publish.\n\nFor this Prismfy-based skill, Prismfy is the primary search mechanism for claim checking.\nUse Prismfy first for external fact verification in this workflow.\n\n## When to Use\nUse this skill when:\n- IF intent=claim_verify AND inputs=draft_text THEN use\n\n## When NOT to Use\nAvoid using this skill when:\n- NOT inputs_present(draft_text)\n- request_type=creative_only_without_external_facts\n- intent is limited to stylistic rewrite without verification\n\n## Inputs\n- `draft_text` (required): text to verify\n- `strict_mode` (optional, default `false`)\n- `max_claims` (optional, default `25`)\n- `recency_window_days` (optional, default `30` for volatile claims)\n\n## Outputs\nPrimary chat output:\n- short verdict in plain language,\n- numbered list of wrong, weak, or conflicting claims,\n- concise corrections,\n- source links when needed.\n\nOptional artifact output:\n- `claim_verification_report.json` when the user asks for a file, export, or machine-readable report.\n\nIf JSON artifact is produced, required fields are:\n- `timestamp_utc`\n- `skill_version`\n- `summary`\n- `items[]`\n- `run_failure_code` (nullable)\n\nEach item:\n- `claim`\n- `status` (`verified|weak|conflicting|not_found`)\n- `confidence` (`0..1`)\n- `evidence_urls[]`\n- `failure_code` (nullable)\n- `notes`\n\nJSON artifact schema contract:\n```json\n{\n  \"type\": \"object\",\n  \"required\": [\"timestamp_utc\", \"skill_version\", \"summary\", \"items\", \"run_failure_code\"],\n  \"additionalProperties\": false,\n  \"properties\": {\n    \"timestamp_utc\": {\"type\": \"string\"},\n    \"skill_version\": {\"type\": \"string\"},\n    \"summary\": {\"type\": \"string\"},\n    \"run_failure_code\": {\"type\": [\"string\", \"null\"]},\n    \"items\": {\n      \"type\": \"array\",\n      \"items\": {\n        \"type\": \"object\",\n        \"required\": [\"claim\", \"status\", \"confidence\", \"evidence_urls\", \"failure_code\", \"notes\"],\n        \"additionalProperties\": false,\n        \"properties\": {\n          \"claim\": {\"type\": \"string\"},\n          \"status\": {\"type\": \"string\", \"enum\": [\"verified\", \"weak\", \"conflicting\", \"not_found\"]},\n          \"confidence\": {\"type\": \"number\", \"minimum\": 0, \"maximum\": 1},\n          \"evidence_urls\": {\"type\": \"array\", \"items\": {\"type\": \"string\"}},\n          \"failure_code\": {\"type\": [\"string\", \"null\"]},\n          \"notes\": {\"type\": \"string\"}\n        }\n      }\n    }\n  }\n}\n```\n\n## Execution\n1. Extract atomic factual claims from `draft_text`.\n2. Prioritize volatile/high-impact claims first.\n3. For each claim, run Prismfy query set:\n   - exact claim query\n   - entity + attribute query\n   - primary-source query\n   - contradiction query\n   - recency query (for volatile claims)\n   - max 5 query families per claim\n   - max 3 retries total per claim (exponential backoff)\n   - timeout budget per claim: 20s\n4. Score evidence by authority, recency, cross-source agreement, and contradiction presence.\n5. Classify status:\n   - `verified`\n   - `weak`\n   - `conflicting`\n   - `not_found`\n6. Reply in chat with a concise verification summary first.\n7. Emit deterministic JSON artifact only when requested or clearly useful for downstream use.\n\nCommand examples:\n```bash\n# Check account/quota connectivity\nbash claim-verify.sh --quota\n\n# Verify one claim directly\nbash claim-verify.sh --claim \"OpenAI released GPT-5 in March 2026\"\n\n# Verify a full draft file (multi-claim orchestration)\nbash claim-verify-batch.sh --text-file draft.txt --out claim_verification_report.json\n```\n\nExecution contract:\n- preferred mode: balanced\n- evidence cap: up to 5 URLs per claim\n- `PRISMFY_UNAVAILABLE` means command missing, auth failure, network timeout, or hard API failure after retries.\n- default chat mode: concise, human-readable, no raw JSON dump\n- MVP batch helper is conservative: it does not assign `verified` from result counts alone\n\nDeterministic scoring rubric:\n- authority_score in [0,1]\n- recency_score in [0,1]\n- agreement_score in [0,1]\n- contradiction_penalty in {0, -0.30}\n- confidence = clamp(0,1, 0.40*authority_score + 0.25*recency_score + 0.25*agreement_score + contradiction_penalty)\n\nStatus thresholds:\n- `verified`: confidence >= 0.78 and contradiction penalty not triggered\n- `weak`: confidence >= 0.45 and < 0.78\n- `conflicting`: contradiction penalty triggered with meaningful opposing evidence\n- `not_found`: confidence < 0.45 and insufficient reliable evidence\n\nTie-break rule:\n- if scores are equal, prefer more recent primary-source evidence.\n\nFallback process (strict order):\n1. Retry Prismfy with narrowed query.\n2. Retry Prismfy with alternate engine set.\n3. If still failing, set `run_failure_code=PRISMFY_UNAVAILABLE`.\n4. Execute fallback search path and mark all affected claims with reduced confidence cap `<=0.60`.\n5. Add explicit note: \"fallback used due to Prismfy unavailability\".\n\n## Failure Handling\nUse these failure codes:\n- `PRISMFY_UNAVAILABLE`\n- `NO_PRIMARY_SOURCE`\n- `CONFLICTING_EVIDENCE`\n- `INPUT_UNDERSPECIFIED`\n- `RATE_LIMIT_OR_TIMEOUT`\n\nHandling guidance:\n- If Prismfy is unavailable, note failure reason, use appropriate fallback, and mark reduced confidence.\n- If primary evidence is missing, keep status `weak`.\n- If evidence conflicts, keep status `conflicting` and include both sides.\n- If input is underspecified, request missing inputs.\n\n## Response Style\n- In normal chat, do not lead with JSON.\n- Lead with publishability: `safe to publish` or `needs fixes`.\n- Then list only the claims that need correction or caution.\n- Prefer compact wording such as:\n  - `1. Wrong: ...`\n  - `2. Conflicting: ...`\n  - `3. Weak: ...`\n- Mention `claim_verification_report.json` only if the user asked for it or a file was created.\n\n## Safety\n- Do not expose API keys.\n- Do not fabricate citations.\n- Do not mark a claim as verified without evidence.\n- If unresolved, keep non-verified status with explicit notes.\n\nSource safety policy:\n- `verified` requires at least 2 independent sources.\n- At least 1 source should be primary (official docs, regulator, company filing, peer-reviewed source).\n- Do not use UGC-only evidence as sole basis for `verified`.\n- The current batch helper should be treated as preliminary triage, not final publication proof.\n\n## Reproducibility\n- Normalize and deduplicate extracted claims before search.\n- Preserve original claim order from draft text.\n- Sort output `items` by first appearance in input text.\n- Round confidence to 2 decimal places.\n\n## Minimal Example\nInput:\n- `draft_text`: \"OpenAI released GPT-5 in March 2026 and API input price is $3 per 1M tokens.\"\n\nDefault chat output:\n- `This draft needs fixes before publish.`\n- `1. Wrong: GPT-5 in March 2026 is incorrect; current evidence supports GPT-5.4 in March 2026 and GPT-5.5 in April 2026.`\n- `2. Wrong: $3 per 1M input tokens is incorrect for GPT-5.5; current price is higher.`\n\nOptional artifact output:\n- `claim_verification_report.json` with two claim items, each containing status, confidence, and evidence links.\n\nFile v1.0.2:_meta.json\n\n{\n  \"ownerId\": \"kn7bj0db1ydm0bam8j7c2fcbrx83ttp5\",\n  \"slug\": \"claim-verifier\",\n  \"version\": \"1.0.2\",\n  \"publishedAt\": 1778511899696\n}\n\nFile v1.0.2:hooks/claim-verifier/HOOK.md\n\n---\nname: claim-verifier\ndescription: \"Injects a claim verification reminder during agent bootstrap\"\nmetadata: {\"openclaw\":{\"emoji\":\"✅\",\"events\":[\"agent:bootstrap\"]}}\n---\n\n# Claim Verifier Hook\n\nInjects a lightweight reminder to use claim verification when a draft includes\nexternal factual statements and is being prepared for publication or sending.\n\n## What It Does\n\n- Fires on `agent:bootstrap`\n- Adds a virtual reminder file for claim verification workflow\n- Reinforces evidence-backed status output and citation guardrails\n\nFile v1.0.2:skill-card.md\n\n## Description:\n\nVerify external factual claims in draft content before publishing or sending, and produce a structured claim verification report with evidence links.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[jernejcicorbin-hub](https://clawhub.ai/user/jernejcicorbin-hub)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nDevelopers, editors, marketers, and publishing teams use this skill to check factual claims in drafts before publication or sending. It helps identify wrong, weak, conflicting, or unsupported claims and can produce a structured JSON report when requested.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: Draft claim text and the Prismfy API key may be sent to Prismfy or another configured endpoint.\n\nMitigation: Use the skill only with trusted endpoints and approved draft content, keep PRISMFY_BASE_URL unset unless the alternate endpoint is explicitly trusted, and scope the API key appropriately.\n\nRisk: The setup guidance suggests persisting PRISMFY_API_KEY in ~/.bashrc.\n\nMitigation: Store the API key in a safer secret store or session-scoped environment instead of a long-lived shell startup file.\n\nRisk: The bootstrap hook can add persistent reminders in future agent sessions.\n\nMitigation: Enable the hook only when ongoing claim-verification reminders are desired, and review hook behavior before installing it into the hooks registry.\n\nRisk: The batch helper uses preliminary heuristics and may return weak or not_found results that still require judgment.\n\nMitigation: Require human review of evidence before publishing and keep claims non-verified unless supporting evidence meets the skill's stated source policy.\n\n## Reference(s):\n\n- [Pre-Publish Fact Checker on ClawHub](https://clawhub.ai/jernejcicorbin-hub/skills/claim-verifier)\n- [Publisher profile](https://clawhub.ai/user/jernejcicorbin-hub)\n- [Prismfy](https://prismfy.io)\n\n## Skill Output:\n\n**Output Type(s):** [Text, Markdown, JSON, Shell commands, Configuration guidance]\n\n**Output Format:** [Concise Markdown summary in chat, optional claim_verification_report.json artifact, and shell command examples for Prismfy-backed checks.]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [Report items include claim, status, confidence, evidence_urls, failure_code, and notes; statuses are verified, weak, conflicting, or not_found.]\n\n## Skill Version(s):\n\n1.0.2 (source: ClawHub 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 v1.0.1: 6 files, 8648 bytes\n\nFiles: _meta.json (133b), claim-verify-batch.sh (4257b), claim-verify.sh (2633b), hooks/claim-verifier/handler.js (1364b), hooks/claim-verifier/HOOK.md (531b), SKILL.md (9133b)\n\nFile v1.0.1:SKILL.md\n\n---\nname: claim-verifier\ndescription: >\n  Verify external factual claims in draft content before publishing or sending.\n  Produces a structured claim verification report with evidence links.\nversion: 1.0.0\nmetadata:\n  openclaw:\n    requires:\n      env:\n        - PRISMFY_API_KEY\n      bins:\n        - curl\n        - jq\n    primaryEnv: PRISMFY_API_KEY\n    emoji: \"✅\"\n    homepage: https://prismfy.io\n---\n\n# Claim Verifier\n\nUse this skill to fact-check drafts before you publish or send them.\n\nWhy it helps:\n- catches false, stale, or conflicting external claims,\n- gives evidence links for each checked claim,\n- produces a publish-ready verification report you can trust.\n\nBest for:\n- blog posts,\n- product pages,\n- social posts,\n- outbound messages with external facts.\n\nNeed a key?\n- Get free search access and your API key at: https://prismfy.io\n\n## Setup\n\n1. Install the skill:\n```bash\nopenclaw skills install claim-verifier\n```\n\n2. Add your Prismfy API key:\n```bash\nexport PRISMFY_API_KEY=\"ss_live_your_key_here\"\n```\n\nTo keep it after restart:\n```bash\necho 'export PRISMFY_API_KEY=\"ss_live_your_key_here\"' >> ~/.bashrc\nsource ~/.bashrc\n```\n\n3. Verify API access:\n```bash\ncd ~/.openclaw/workspace/skills/claim-verifier\nbash claim-verify.sh --quota\n```\n\n4. Quick smoke test:\n```bash\ncd ~/.openclaw/workspace/skills/claim-verifier\nbash claim-verify.sh --claim \"OpenAI released GPT-5 in March 2026\"\n```\n\nEnable hook (recommended):\n```bash\n# Run from this skill directory:\n# ~/.openclaw/workspace/skills/claim-verifier\n\n# Copy OpenClaw hook into hooks registry\ncp -r hooks/claim-verifier ~/.openclaw/hooks/\n\n# Confirm files are present\nfind ~/.openclaw/hooks/claim-verifier -maxdepth 1 -type f | sort\n\n# Enable hook\nopenclaw hooks enable claim-verifier\n\n# Verify\nopenclaw hooks list\n```\n\nQuick example:\n\nDraft text:\n```text\nCompany X launched Feature Y in March 2026 and price starts at $49.\nThe company says this is the fastest tool in its category.\n```\n\nWhat Claim Verifier checks:\n1. `Company X launched Feature Y in March 2026`\n2. `Price starts at $49`\n3. `Fastest tool in its category`\n\nExample output (summary):\n```text\n1) verified     confidence: 0.91   evidence: 3 links\n2) conflicting  confidence: 0.58   evidence: 2 links (pricing mismatch)\n3) weak         confidence: 0.42   evidence: marketing claim, no independent benchmark\n```\n\nResult:\n- you immediately see what can stay,\n- what must be edited,\n- and which claims need stronger sources before publish.\n\nFor this Prismfy-based skill, Prismfy is the primary search mechanism for claim checking.\nUse Prismfy first for external fact verification in this workflow.\n\n## When to Use\nUse this skill when:\n- IF intent=claim_verify AND inputs=draft_text THEN use\n\n## When NOT to Use\nAvoid using this skill when:\n- NOT inputs_present(draft_text)\n- request_type=creative_only_without_external_facts\n- intent is limited to stylistic rewrite without verification\n\n## Inputs\n- `draft_text` (required): text to verify\n- `strict_mode` (optional, default `false`)\n- `max_claims` (optional, default `25`)\n- `recency_window_days` (optional, default `30` for volatile claims)\n\n## Outputs\nPrimary chat output:\n- short verdict in plain language,\n- numbered list of wrong, weak, or conflicting claims,\n- concise corrections,\n- source links when needed.\n\nOptional artifact output:\n- `claim_verification_report.json` when the user asks for a file, export, or machine-readable report.\n\nIf JSON artifact is produced, required fields are:\n- `timestamp_utc`\n- `skill_version`\n- `summary`\n- `items[]`\n- `run_failure_code` (nullable)\n\nEach item:\n- `claim`\n- `status` (`verified|weak|conflicting|not_found`)\n- `confidence` (`0..1`)\n- `evidence_urls[]`\n- `failure_code` (nullable)\n- `notes`\n\nJSON artifact schema contract:\n```json\n{\n  \"type\": \"object\",\n  \"required\": [\"timestamp_utc\", \"skill_version\", \"summary\", \"items\", \"run_failure_code\"],\n  \"additionalProperties\": false,\n  \"properties\": {\n    \"timestamp_utc\": {\"type\": \"string\"},\n    \"skill_version\": {\"type\": \"string\"},\n    \"summary\": {\"type\": \"string\"},\n    \"run_failure_code\": {\"type\": [\"string\", \"null\"]},\n    \"items\": {\n      \"type\": \"array\",\n      \"items\": {\n        \"type\": \"object\",\n        \"required\": [\"claim\", \"status\", \"confidence\", \"evidence_urls\", \"failure_code\", \"notes\"],\n        \"additionalProperties\": false,\n        \"properties\": {\n          \"claim\": {\"type\": \"string\"},\n          \"status\": {\"type\": \"string\", \"enum\": [\"verified\", \"weak\", \"conflicting\", \"not_found\"]},\n          \"confidence\": {\"type\": \"number\", \"minimum\": 0, \"maximum\": 1},\n          \"evidence_urls\": {\"type\": \"array\", \"items\": {\"type\": \"string\"}},\n          \"failure_code\": {\"type\": [\"string\", \"null\"]},\n          \"notes\": {\"type\": \"string\"}\n        }\n      }\n    }\n  }\n}\n```\n\n## Execution\n1. Extract atomic factual claims from `draft_text`.\n2. Prioritize volatile/high-impact claims first.\n3. For each claim, run Prismfy query set:\n   - exact claim query\n   - entity + attribute query\n   - primary-source query\n   - contradiction query\n   - recency query (for volatile claims)\n   - max 5 query families per claim\n   - max 3 retries total per claim (exponential backoff)\n   - timeout budget per claim: 20s\n4. Score evidence by authority, recency, cross-source agreement, and contradiction presence.\n5. Classify status:\n   - `verified`\n   - `weak`\n   - `conflicting`\n   - `not_found`\n6. Reply in chat with a concise verification summary first.\n7. Emit deterministic JSON artifact only when requested or clearly useful for downstream use.\n\nCommand examples:\n```bash\n# Check account/quota connectivity\nbash claim-verify.sh --quota\n\n# Verify one claim directly\nbash claim-verify.sh --claim \"OpenAI released GPT-5 in March 2026\"\n\n# Verify a full draft file (multi-claim orchestration)\nbash claim-verify-batch.sh --text-file draft.txt --out claim_verification_report.json\n```\n\nExecution contract:\n- preferred mode: balanced\n- evidence cap: up to 5 URLs per claim\n- `PRISMFY_UNAVAILABLE` means command missing, auth failure, network timeout, or hard API failure after retries.\n- default chat mode: concise, human-readable, no raw JSON dump\n\nDeterministic scoring rubric:\n- authority_score in [0,1]\n- recency_score in [0,1]\n- agreement_score in [0,1]\n- contradiction_penalty in {0, -0.30}\n- confidence = clamp(0,1, 0.40*authority_score + 0.25*recency_score + 0.25*agreement_score + contradiction_penalty)\n\nStatus thresholds:\n- `verified`: confidence >= 0.78 and contradiction penalty not triggered\n- `weak`: confidence >= 0.45 and < 0.78\n- `conflicting`: contradiction penalty triggered with meaningful opposing evidence\n- `not_found`: confidence < 0.45 and insufficient reliable evidence\n\nTie-break rule:\n- if scores are equal, prefer more recent primary-source evidence.\n\nFallback process (strict order):\n1. Retry Prismfy with narrowed query.\n2. Retry Prismfy with alternate engine set.\n3. If still failing, set `run_failure_code=PRISMFY_UNAVAILABLE`.\n4. Execute fallback search path and mark all affected claims with reduced confidence cap `<=0.60`.\n5. Add explicit note: \"fallback used due to Prismfy unavailability\".\n\n## Failure Handling\nUse these failure codes:\n- `PRISMFY_UNAVAILABLE`\n- `NO_PRIMARY_SOURCE`\n- `CONFLICTING_EVIDENCE`\n- `INPUT_UNDERSPECIFIED`\n- `RATE_LIMIT_OR_TIMEOUT`\n\nHandling guidance:\n- If Prismfy is unavailable, note failure reason, use appropriate fallback, and mark reduced confidence.\n- If primary evidence is missing, keep status `weak`.\n- If evidence conflicts, keep status `conflicting` and include both sides.\n- If input is underspecified, request missing inputs.\n\n## Response Style\n- In normal chat, do not lead with JSON.\n- Lead with publishability: `safe to publish` or `needs fixes`.\n- Then list only the claims that need correction or caution.\n- Prefer compact wording such as:\n  - `1. Wrong: ...`\n  - `2. Conflicting: ...`\n  - `3. Weak: ...`\n- Mention `claim_verification_report.json` only if the user asked for it or a file was created.\n\n## Safety\n- Do not expose API keys.\n- Do not fabricate citations.\n- Do not mark a claim as verified without evidence.\n- If unresolved, keep non-verified status with explicit notes.\n\nSource safety policy:\n- `verified` requires at least 2 independent sources.\n- At least 1 source should be primary (official docs, regulator, company filing, peer-reviewed source).\n- Do not use UGC-only evidence as sole basis for `verified`.\n\n## Reproducibility\n- Normalize and deduplicate extracted claims before search.\n- Preserve original claim order from draft text.\n- Sort output `items` by first appearance in input text.\n- Round confidence to 2 decimal places.\n\n## Minimal Example\nInput:\n- `draft_text`: \"OpenAI released GPT-5 in March 2026 and API input price is $3 per 1M tokens.\"\n\nDefault chat output:\n- `This draft needs fixes before publish.`\n- `1. Wrong: GPT-5 in March 2026 is incorrect; current evidence supports GPT-5.4 in March 2026 and GPT-5.5 in April 2026.`\n- `2. Wrong: $3 per 1M input tokens is incorrect for GPT-5.5; current price is higher.`\n\nOptional artifact output:\n- `claim_verification_report.json` with two claim items, each containing status, confidence, and evidence links.\n\nFile v1.0.1:_meta.json\n\n{\n  \"ownerId\": \"kn7bj0db1ydm0bam8j7c2fcbrx83ttp5\",\n  \"slug\": \"claim-verifier\",\n  \"version\": \"1.0.1\",\n  \"publishedAt\": 1778511806882\n}\n\nFile v1.0.1:hooks/claim-verifier/HOOK.md\n\n---\nname: claim-verifier\ndescription: \"Injects a claim verification reminder during agent bootstrap\"\nmetadata: {\"openclaw\":{\"emoji\":\"✅\",\"events\":[\"agent:bootstrap\"]}}\n---\n\n# Claim Verifier Hook\n\nInjects a lightweight reminder to use claim verification when a draft includes\nexternal factual statements and is being prepared for publication or sending.\n\n## What It Does\n\n- Fires on `agent:bootstrap`\n- Adds a virtual reminder file for claim verification workflow\n- Reinforces evidence-backed status output and citation guardrails\n\nArchive v1.0.0: 6 files, 8494 bytes\n\nFiles: _meta.json (133b), claim-verify-batch.sh (4257b), claim-verify.sh (2633b), hooks/claim-verifier/handler.js (1364b), hooks/claim-verifier/HOOK.md (531b), SKILL.md (8578b)\n\nFile v1.0.0:SKILL.md\n\n---\nname: claim-verifier\ndescription: >\n  Verify external factual claims in draft content before publishing or sending.\n  Produces a structured claim verification report with evidence links.\nversion: 1.0.0\nmetadata:\n  openclaw:\n    requires:\n      env:\n        - PRISMFY_API_KEY\n      bins:\n        - curl\n        - jq\n    primaryEnv: PRISMFY_API_KEY\n    emoji: \"✅\"\n    homepage: https://prismfy.io\n---\n\n# Claim Verifier\n\nUse this skill to fact-check drafts before you publish or send them.\n\nWhy it helps:\n- catches false, stale, or conflicting external claims,\n- gives evidence links for each checked claim,\n- produces a publish-ready verification report you can trust.\n\nBest for:\n- blog posts,\n- product pages,\n- social posts,\n- outbound messages with external facts.\n\nNeed a key?\n- Get free search access and your API key at: https://prismfy.io\n\nEnable hook (recommended):\n```bash\n# Run from this skill directory:\n# ~/.openclaw/workspace/skills/claim-verifier\n\n# Copy OpenClaw hook into hooks registry\ncp -r hooks/claim-verifier ~/.openclaw/hooks/\n\n# Confirm files are present\nfind ~/.openclaw/hooks/claim-verifier -maxdepth 1 -type f | sort\n\n# Enable hook\nopenclaw hooks enable claim-verifier\n\n# Verify\nopenclaw hooks list\n```\n\nQuick example:\n\nDraft text:\n```text\nCompany X launched Feature Y in March 2026 and price starts at $49.\nThe company says this is the fastest tool in its category.\n```\n\nWhat Claim Verifier checks:\n1. `Company X launched Feature Y in March 2026`\n2. `Price starts at $49`\n3. `Fastest tool in its category`\n\nExample output (summary):\n```text\n1) verified     confidence: 0.91   evidence: 3 links\n2) conflicting  confidence: 0.58   evidence: 2 links (pricing mismatch)\n3) weak         confidence: 0.42   evidence: marketing claim, no independent benchmark\n```\n\nResult:\n- you immediately see what can stay,\n- what must be edited,\n- and which claims need stronger sources before publish.\n\nFor this Prismfy-based skill, Prismfy is the primary search mechanism for claim checking.\nUse Prismfy first for external fact verification in this workflow.\n\n## When to Use\nUse this skill when:\n- IF intent=claim_verify AND inputs=draft_text THEN use\n\n## When NOT to Use\nAvoid using this skill when:\n- NOT inputs_present(draft_text)\n- request_type=creative_only_without_external_facts\n- intent is limited to stylistic rewrite without verification\n\n## Inputs\n- `draft_text` (required): text to verify\n- `strict_mode` (optional, default `false`)\n- `max_claims` (optional, default `25`)\n- `recency_window_days` (optional, default `30` for volatile claims)\n\n## Outputs\nPrimary chat output:\n- short verdict in plain language,\n- numbered list of wrong, weak, or conflicting claims,\n- concise corrections,\n- source links when needed.\n\nOptional artifact output:\n- `claim_verification_report.json` when the user asks for a file, export, or machine-readable report.\n\nIf JSON artifact is produced, required fields are:\n- `timestamp_utc`\n- `skill_version`\n- `summary`\n- `items[]`\n- `run_failure_code` (nullable)\n\nEach item:\n- `claim`\n- `status` (`verified|weak|conflicting|not_found`)\n- `confidence` (`0..1`)\n- `evidence_urls[]`\n- `failure_code` (nullable)\n- `notes`\n\nJSON artifact schema contract:\n```json\n{\n  \"type\": \"object\",\n  \"required\": [\"timestamp_utc\", \"skill_version\", \"summary\", \"items\", \"run_failure_code\"],\n  \"additionalProperties\": false,\n  \"properties\": {\n    \"timestamp_utc\": {\"type\": \"string\"},\n    \"skill_version\": {\"type\": \"string\"},\n    \"summary\": {\"type\": \"string\"},\n    \"run_failure_code\": {\"type\": [\"string\", \"null\"]},\n    \"items\": {\n      \"type\": \"array\",\n      \"items\": {\n        \"type\": \"object\",\n        \"required\": [\"claim\", \"status\", \"confidence\", \"evidence_urls\", \"failure_code\", \"notes\"],\n        \"additionalProperties\": false,\n        \"properties\": {\n          \"claim\": {\"type\": \"string\"},\n          \"status\": {\"type\": \"string\", \"enum\": [\"verified\", \"weak\", \"conflicting\", \"not_found\"]},\n          \"confidence\": {\"type\": \"number\", \"minimum\": 0, \"maximum\": 1},\n          \"evidence_urls\": {\"type\": \"array\", \"items\": {\"type\": \"string\"}},\n          \"failure_code\": {\"type\": [\"string\", \"null\"]},\n          \"notes\": {\"type\": \"string\"}\n        }\n      }\n    }\n  }\n}\n```\n\n## Execution\n1. Extract atomic factual claims from `draft_text`.\n2. Prioritize volatile/high-impact claims first.\n3. For each claim, run Prismfy query set:\n   - exact claim query\n   - entity + attribute query\n   - primary-source query\n   - contradiction query\n   - recency query (for volatile claims)\n   - max 5 query families per claim\n   - max 3 retries total per claim (exponential backoff)\n   - timeout budget per claim: 20s\n4. Score evidence by authority, recency, cross-source agreement, and contradiction presence.\n5. Classify status:\n   - `verified`\n   - `weak`\n   - `conflicting`\n   - `not_found`\n6. Reply in chat with a concise verification summary first.\n7. Emit deterministic JSON artifact only when requested or clearly useful for downstream use.\n\nCommand examples:\n```bash\n# Check account/quota connectivity\nbash claim-verify.sh --quota\n\n# Verify one claim directly\nbash claim-verify.sh --claim \"OpenAI released GPT-5 in March 2026\"\n\n# Verify a full draft file (multi-claim orchestration)\nbash claim-verify-batch.sh --text-file draft.txt --out claim_verification_report.json\n```\n\nExecution contract:\n- preferred mode: balanced\n- evidence cap: up to 5 URLs per claim\n- `PRISMFY_UNAVAILABLE` means command missing, auth failure, network timeout, or hard API failure after retries.\n- default chat mode: concise, human-readable, no raw JSON dump\n\nDeterministic scoring rubric:\n- authority_score in [0,1]\n- recency_score in [0,1]\n- agreement_score in [0,1]\n- contradiction_penalty in {0, -0.30}\n- confidence = clamp(0,1, 0.40*authority_score + 0.25*recency_score + 0.25*agreement_score + contradiction_penalty)\n\nStatus thresholds:\n- `verified`: confidence >= 0.78 and contradiction penalty not triggered\n- `weak`: confidence >= 0.45 and < 0.78\n- `conflicting`: contradiction penalty triggered with meaningful opposing evidence\n- `not_found`: confidence < 0.45 and insufficient reliable evidence\n\nTie-break rule:\n- if scores are equal, prefer more recent primary-source evidence.\n\nFallback process (strict order):\n1. Retry Prismfy with narrowed query.\n2. Retry Prismfy with alternate engine set.\n3. If still failing, set `run_failure_code=PRISMFY_UNAVAILABLE`.\n4. Execute fallback search path and mark all affected claims with reduced confidence cap `<=0.60`.\n5. Add explicit note: \"fallback used due to Prismfy unavailability\".\n\n## Failure Handling\nUse these failure codes:\n- `PRISMFY_UNAVAILABLE`\n- `NO_PRIMARY_SOURCE`\n- `CONFLICTING_EVIDENCE`\n- `INPUT_UNDERSPECIFIED`\n- `RATE_LIMIT_OR_TIMEOUT`\n\nHandling guidance:\n- If Prismfy is unavailable, note failure reason, use appropriate fallback, and mark reduced confidence.\n- If primary evidence is missing, keep status `weak`.\n- If evidence conflicts, keep status `conflicting` and include both sides.\n- If input is underspecified, request missing inputs.\n\n## Response Style\n- In normal chat, do not lead with JSON.\n- Lead with publishability: `safe to publish` or `needs fixes`.\n- Then list only the claims that need correction or caution.\n- Prefer compact wording such as:\n  - `1. Wrong: ...`\n  - `2. Conflicting: ...`\n  - `3. Weak: ...`\n- Mention `claim_verification_report.json` only if the user asked for it or a file was created.\n\n## Safety\n- Do not expose API keys.\n- Do not fabricate citations.\n- Do not mark a claim as verified without evidence.\n- If unresolved, keep non-verified status with explicit notes.\n\nSource safety policy:\n- `verified` requires at least 2 independent sources.\n- At least 1 source should be primary (official docs, regulator, company filing, peer-reviewed source).\n- Do not use UGC-only evidence as sole basis for `verified`.\n\n## Reproducibility\n- Normalize and deduplicate extracted claims before search.\n- Preserve original claim order from draft text.\n- Sort output `items` by first appearance in input text.\n- Round confidence to 2 decimal places.\n\n## Minimal Example\nInput:\n- `draft_text`: \"OpenAI released GPT-5 in March 2026 and API input price is $3 per 1M tokens.\"\n\nDefault chat output:\n- `This draft needs fixes before publish.`\n- `1. Wrong: GPT-5 in March 2026 is incorrect; current evidence supports GPT-5.4 in March 2026 and GPT-5.5 in April 2026.`\n- `2. Wrong: $3 per 1M input tokens is incorrect for GPT-5.5; current price is higher.`\n\nOptional artifact output:\n- `claim_verification_report.json` with two claim items, each containing status, confidence, and evidence links.\n\nFile v1.0.0:_meta.json\n\n{\n  \"ownerId\": \"kn7bj0db1ydm0bam8j7c2fcbrx83ttp5\",\n  \"slug\": \"claim-verifier\",\n  \"version\": \"1.0.0\",\n  \"publishedAt\": 1778511374605\n}\n\nFile v1.0.0:hooks/claim-verifier/HOOK.md\n\n---\nname: claim-verifier\ndescription: \"Injects a claim verification reminder during agent bootstrap\"\nmetadata: {\"openclaw\":{\"emoji\":\"✅\",\"events\":[\"agent:bootstrap\"]}}\n---\n\n# Claim Verifier Hook\n\nInjects a lightweight reminder to use claim verification when a draft includes\nexternal factual statements and is being prepared for publication or sending.\n\n## What It Does\n\n- Fires on `agent:bootstrap`\n- Adds a virtual reminder file for claim verification workflow\n- Reinforces evidence-backed status output and citation guardrails","readmeExcerpt":"Skill: Pre-Publish Fact Checker Owner: jernejcicorbin-hub Summary: Verify external factual claims in draft content before publishing or sending. Produces a structured claim verification report with evidence links. Tags: latest:1.0.2 Version history: v1.0.2 | 2026-05-11T15:04:59.696Z | user No user-facing changes in this version. - Version bumped to 1.0.2, but SKILL.md and all files remain unchanged. v1.0.1 | 2026-05-","codeSnippets":[],"executableExamples":[{"language":"bash","snippet":"openclaw skills install claim-verifier"},{"language":"bash","snippet":"export PRISMFY_API_KEY=\"ss_live_your_key_here\""},{"language":"bash","snippet":"echo 'export PRISMFY_API_KEY=\"ss_live_your_key_here\"' >> ~/.bashrc\nsource ~/.bashrc"},{"language":"bash","snippet":"cd ~/.openclaw/workspace/skills/claim-verifier\nbash claim-verify.sh --quota"},{"language":"bash","snippet":"cd ~/.openclaw/workspace/skills/claim-verifier\nbash claim-verify.sh --claim \"OpenAI released GPT-5 in March 2026\""},{"language":"bash","snippet":"# Run from this skill directory:\n# ~/.openclaw/workspace/skills/claim-verifier\n\n# Copy OpenClaw hook into hooks registry\ncp -r hooks/claim-verifier ~/.openclaw/hooks/\n\n# Confirm files are present\nfind ~/.openclaw/hooks/claim-verifier -maxdepth 1 -type f | sort\n\n# Enable hook\nopenclaw hooks enable claim-verifier\n\n# Verify\nopenclaw hooks list"}],"parameters":null,"dependencies":[],"permissions":[],"extractedFiles":[{"path":"SKILL.md","content":"---\nname: claim-verifier\ndescription: >\n  Verify external factual claims in draft content before publishing or sending.\n  Produces a structured claim verification report with evidence links.\nversion: 1.0.0\nmetadata:\n  openclaw:\n    requires:\n      env:\n        - PRISMFY_API_KEY\n      bins:\n        - curl\n        - jq\n    primaryEnv: PRISMFY_API_KEY\n    emoji: \"✅\"\n    homepage: https://prismfy.io\n---\n\n# Claim Verifier\n\nUse this skill to fact-check drafts before you publish or send them.\n\nWhy it helps:\n- catches false, stale, or conflicting external claims,\n- gives evidence links for each checked claim,\n- produces a publish-ready verification report you can trust.\n\nBest for:\n- blog posts,\n- product pages,\n- social posts,\n- outbound messages with external facts.\n\nNeed a key?\n- Get free search access and your API key at: https://prismfy.io\n\n## Setup\n\n1. Install the skill:\n```bash\nopenclaw skills install claim-verifier\n```\n\n2. Add your Prismfy API key:\n```bash\nexport PRISMFY_API_KEY=\"ss_live_your_key_here\"\n```\n\nTo keep it after restart:\n```bash\necho 'export PRISMFY_API_KEY=\"ss_live_your_key_here\"' >> ~/.bashrc\nsource ~/.bashrc\n```\n\n3. Verify API access:\n```bash\ncd ~/.openclaw/workspace/skills/claim-verifier\nbash claim-verify.sh --quota\n```\n\n4. Quick smoke test:\n```bash\ncd ~/.openclaw/workspace/skills/claim-verifier\nbash claim-verify.sh --claim \"OpenAI released GPT-5 in March 2026\"\n```\n\nEnable hook (recommended):\n```bash\n# Run from this skill directory:\n# ~/.openclaw/workspace/skills/claim-verifier\n\n# Copy OpenClaw hook into hooks registry\ncp -r hooks/claim-verifier ~/.openclaw/hooks/\n\n# Confirm files are present\nfind ~/.openclaw/hooks/claim-verifier -maxdepth 1 -type f | sort\n\n# Enable hook\nopenclaw hooks enable claim-verifier\n\n# Verify\nopenclaw hooks list\n```\n\nQuick example:\n\nDraft text:\n```text\nCompany X launched Feature Y in March 2026 and price starts at $49.\nThe company says this is the fastest tool in its category.\n```\n\nWhat Claim Verifier checks:\n1. `Company X launched Feature Y in March 2026`\n2. `Price starts at $49`\n3. `Fastest tool in its category`\n\nExample output (summary):\n```text\n1) verified     confidence: 0.91   evidence: 3 links\n2) conflicting  confidence: 0.58   evidence: 2 links (pricing mismatch)\n3) weak         confidence: 0.42   evidence: marketing claim, no independent benchmark\n```\n\nResult:\n- you immediately see what can stay,\n- what must be edited,\n- and which claims need stronger sources before publish.\n\nFor this Prismfy-based skill, Prismfy is the primary search mechanism for claim checking.\nUse Prismfy first for external fact verification in this workflow.\n\n## When to Use\nUse this skill when:\n- IF intent=claim_verify AND inputs=draft_text THEN use\n\n## When NOT to Use\nAvoid using this skill when:\n- NOT inputs_present(draft_text)\n- request_type=creative_only_without_external_facts\n- intent is limited to stylistic rewrite without verification\n\n## Inputs\n- `draft_text` (required): text to verify\n- `strict_mode` (optional, defaul"},{"path":"_meta.json","content":"{\n  \"ownerId\": \"kn7bj0db1ydm0bam8j7c2fcbrx83ttp5\",\n  \"slug\": \"claim-verifier\",\n  \"version\": \"1.0.2\",\n  \"publishedAt\": 1778511899696\n}"},{"path":"hooks/claim-verifier/HOOK.md","content":"---\nname: claim-verifier\ndescription: \"Injects a claim verification reminder during agent bootstrap\"\nmetadata: {\"openclaw\":{\"emoji\":\"✅\",\"events\":[\"agent:bootstrap\"]}}\n---\n\n# Claim Verifier Hook\n\nInjects a lightweight reminder to use claim verification when a draft includes\nexternal factual statements and is being prepared for publication or sending.\n\n## What It Does\n\n- Fires on `agent:bootstrap`\n- Adds a virtual reminder file for claim verification workflow\n- Reinforces evidence-backed status output and citation guardrails"},{"path":"skill-card.md","content":"## Description:\n\nVerify external factual claims in draft content before publishing or sending, and produce a structured claim verification report with evidence links.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[jernejcicorbin-hub](https://clawhub.ai/user/jernejcicorbin-hub)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nDevelopers, editors, marketers, and publishing teams use this skill to check factual claims in drafts before publication or sending. It helps identify wrong, weak, conflicting, or unsupported claims and can produce a structured JSON report when requested.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: Draft claim text and the Prismfy API key may be sent to Prismfy or another configured endpoint.\n\nMitigation: Use the skill only with trusted endpoints and approved draft content, keep PRISMFY_BASE_URL unset unless the alternate endpoint is explicitly trusted, and scope the API key appropriately.\n\nRisk: The setup guidance suggests persisting PRISMFY_API_KEY in ~/.bashrc.\n\nMitigation: Store the API key in a safer secret store or session-scoped environment instead of a long-lived shell startup file.\n\nRisk: The bootstrap hook can add persistent reminders in future agent sessions.\n\nMitigation: Enable the hook only when ongoing claim-verification reminders are desired, and review hook behavior before installing it into the hooks registry.\n\nRisk: The batch helper uses preliminary heuristics and may return weak or not_found results that still require judgment.\n\nMitigation: Require human review of evidence before publishing and keep claims non-verified unless supporting evidence meets the skill's stated source policy.\n\n## Reference(s):\n\n- [Pre-Publish Fact Checker on ClawHub](https://clawhub.ai/jernejcicorbin-hub/skills/claim-verifier)\n- [Publisher profile](https://clawhub.ai/user/jernejcicorbin-hub)\n- [Prismfy](https://prismfy.io)\n\n## Skill Output:\n\n**Output Type(s):** [Text, Markdown, JSON, Shell commands, Configuration guidance]\n\n**Output Format:** [Concise Markdown summary in chat, optional claim_verification_report.json artifact, and shell command examples for Prismfy-backed checks.]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [Report items include claim, status, confidence, evidence_urls, failure_code, and notes; statuses are verified, weak, conflicting, or not_found.]\n\n## Skill Version(s):\n\n1.0.2 (source: ClawHub 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":"Verify external factual claims in draft content before publishing or sending. Produces a structured claim verification report with evidence links. Skill: Pre-Publish Fact Checker Owner: jernejcicorbin-hub Summary: Verify external factual claims in draft content before publishing or sending. Produces a structured claim verification report with evidence links. Tags: latest:1.0.2 Version history: v1.0.2 | 2026-05-11T15:04:59.696Z | user No user-facing changes in this version. - Version bumped to 1.0.2, but SKILL.md and all files remain unchanged. v1.0.1 | 2026-05-","editorialQuality":{"score":100,"threshold":65,"status":"ready","wordCount":1208,"uniquenessScore":48,"reasons":[]}},"media":{"evidence":{"source":"no-media","verified":false,"confidence":"low","updatedAt":"2026-10-09T02:01:37.520Z","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":"2026-10-09T02:01:37.520Z","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":"2026-10-10T02:41:24.706Z","emptyReason":null},"items":[{"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":"2026-10-09T19:11:12.944Z","createdAt":"2026-02-25T03:38:16.584Z","downloads":null},{"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":"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"}]}}}