{"id":"25fba2ef-f7af-4078-8c93-44d6605c2d3a","entityType":"agent","slug":"clawhub-christianye-engineering-discipline","name":"Engineering Discipline","canonicalUrl":"https://www.xpersona.co/agent/clawhub-christianye-engineering-discipline","canonicalPath":"/agent/clawhub-christianye-engineering-discipline","generatedAt":"2026-10-11T20:58:03.493Z","source":"CLAWHUB","claimStatus":"UNCLAIMED","verificationTier":"NONE","summary":{"evidence":{"source":"editorial-content","verified":true,"confidence":"high","updatedAt":"2026-10-11T17:29:07.734Z","emptyReason":null},"description":"Production-grade AI coding discipline. Prevents the top 4 AI coding sins: acting without thinking, over-engineering, collateral damage, and vague execution.... Skill: Engineering Discipline Owner: christianye Summary: Production-grade AI coding discipline. Prevents the top 4 AI coding sins: acting without thinking, over-engineering, collateral damage, and vague execution.... Tags: latest:1.0.1 Version history: v1.0.1 | 2026-06-10T15:55:55.173Z | user Expand description with Chinese trigger keywords for better discoverability (反合理化 / 三层一致性检查 / Karpathy 四规则 / 开发纪律 etc.). No c","descriptionLabel":"Technical summary","evidenceSummary":"Capability contract not published. No trust telemetry is available yet. 1K downloads reported by the source. Last updated 10/11/2026.","installCommand":"clawhub skill install s17bsjyfzzr9b73vsdr9mcdk9n84f96d:engineering-discipline","sourceUrl":"https://clawhub.ai/christianye/engineering-discipline","homepage":"https://clawhub.ai/christianye/skills/engineering-discipline","primaryLinks":[{"label":"View on ClawHub","url":"https://clawhub.ai/christianye/engineering-discipline","kind":"source"},{"label":"Homepage","url":"https://clawhub.ai/christianye/skills/engineering-discipline","kind":"homepage"}],"safetyScore":84,"overallRank":62,"popularityScore":60,"trustScore":null,"claimedByName":null,"isOwner":false,"seoDescription":"Production-grade AI coding discipline. Prevents the top 4 AI coding sins: acting without thinking, over-engineering, collateral damage, and vague execution.... "},"coverage":{"evidence":{"source":"public-profile","verified":false,"confidence":"medium","updatedAt":"2026-10-11T17:29:07.734Z","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":"2026-10-11T17:29:07.734Z","emptyReason":null},"stars":null,"forks":null,"downloads":1018,"packageName":null,"latestVersion":"1.0.1","tractionLabel":"1K downloads"},"release":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-11T17:29:07.722Z","emptyReason":null},"lastUpdatedAt":"2026-10-11T17:29:07.734Z","lastCrawledAt":"2026-10-11T17:29:07.722Z","lastIndexedAt":null,"nextCrawlAt":"2026-10-12T17:29:07.722Z","lastVerifiedAt":null,"highlights":[{"version":"1.0.1","createdAt":"2026-06-10T15:55:55.173Z","changelog":"Expand description with Chinese trigger keywords for better discoverability (反合理化 / 三层一致性检查 / Karpathy 四规则 / 开发纪律 etc.). No content changes.","fileCount":3,"zipByteSize":4667},{"version":"1.0.0","createdAt":"2026-04-24T09:40:54.273Z","changelog":"Initial release introducing a standardized set of coding discipline rules for AI coding assistants: - Codifies Karpathy's 4 foundational rules for AI code generation (think before coding, simplicity first, surgical changes, goal-driven execution). - Adds advanced safety measures: 3-layer consistency checks, anti-rationalization habits, verification loops, pre-change snapshots, and context hygiene guidelines. - Provides clear, actionable behaviors to prevent common AI coding errors like overengineering, collateral changes, and vague fixes. - Designed for seamless integration with Claude Code, Cursor, Copilot, and other AI coding tools. - Intended as a persistent behavior layer for all coding sessions, especially in production-grade environments.","fileCount":3,"zipByteSize":4425}]},"execution":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No published capability contract is available yet."},"installCommand":"clawhub skill install s17bsjyfzzr9b73vsdr9mcdk9n84f96d:engineering-discipline","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-christianye-engineering-discipline/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-christianye-engineering-discipline/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-christianye-engineering-discipline/trust"},"curlExamples":["curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-christianye-engineering-discipline/snapshot\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-christianye-engineering-discipline/contract\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-christianye-engineering-discipline/trust\""],"jsonRequestTemplate":{"query":"summarize this repo","constraints":{"maxLatencyMs":2000,"protocolPreference":["OPENCLEW"]}},"jsonResponseTemplate":{"ok":true,"result":{"summary":"...","confidence":0.9},"meta":{"source":"CLAWHUB","generatedAt":"2026-10-11T20:58:03.493Z"}},"retryPolicy":{"maxAttempts":3,"backoffMs":[500,1500,3500],"retryableConditions":["HTTP_429","HTTP_503","NETWORK_TIMEOUT"]}},"endpoints":{"dossierUrl":"https://www.xpersona.co/api/v1/agents/clawhub-christianye-engineering-discipline/dossier","snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-christianye-engineering-discipline/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-christianye-engineering-discipline/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-christianye-engineering-discipline/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":"2026-10-11T17:29:07.734Z","emptyReason":null},"readme":"Skill: Engineering Discipline\n\nOwner: christianye\n\nSummary: Production-grade AI coding discipline. Prevents the top 4 AI coding sins: acting without thinking, over-engineering, collateral damage, and vague execution....\n\nTags: latest:1.0.1\n\nVersion history:\n\nv1.0.1 | 2026-06-10T15:55:55.173Z | user\n\nExpand description with Chinese trigger keywords for better discoverability (反合理化 / 三层一致性检查 / Karpathy 四规则 / 开发纪律 etc.). No content changes.\n\nv1.0.0 | 2026-04-24T09:40:54.273Z | auto\n\nInitial release introducing a standardized set of coding discipline rules for AI coding assistants:\n\n- Codifies Karpathy's 4 foundational rules for AI code generation (think before coding, simplicity first, surgical changes, goal-driven execution).\n- Adds advanced safety measures: 3-layer consistency checks, anti-rationalization habits, verification loops, pre-change snapshots, and context hygiene guidelines.\n- Provides clear, actionable behaviors to prevent common AI coding errors like overengineering, collateral changes, and vague fixes.\n- Designed for seamless integration with Claude Code, Cursor, Copilot, and other AI coding tools.\n- Intended as a persistent behavior layer for all coding sessions, especially in production-grade environments.\n\nArchive index:\n\nArchive v1.0.1: 3 files, 4667 bytes\n\nFiles: skill-card.md (2124b), SKILL.md (6059b), _meta.json (141b)\n\nFile v1.0.1:SKILL.md\n\n---\nname: Engineering Discipline\nslug: engineering-discipline\nversion: 1.0.1\ndescription: \"Production-grade AI coding discipline. Prevents the top 4 AI coding sins: acting without thinking, over-engineering, collateral damage, and vague execution. Triggers: 'engineering discipline', '过度工程', '反合理化', '三层一致性检查', 'AI coding rules', 'Karpathy 四规则', '写代码前先想', '手术刀修复'. Use when starting any coding task, before large refactors, or when catching yourself rationalizing shortcuts. Based on Karpathy's 4 rules + battle-tested additions (3-layer consistency checks, anti-rationalization, verification loops, surgical diffs). Works with Claude Code, Cursor, Copilot, OpenClaw, and any AI coding assistant. Triggers: 'engineering checklist'、'coding discipline'、'production quality'、'AI coding guardrails'、'开发纪律'。\"\nhomepage: https://clawhub.ai/skills/engineering-discipline\nmetadata: {\"clawdbot\":{\"emoji\":\"🔧\",\"os\":[\"linux\",\"darwin\",\"win32\"]}}\n---\n\n## When to Use\n\nApply this skill to **every coding session**. It's not a task-specific tool — it's a permanent behavior modifier for AI coding assistants.\n\nEspecially critical when:\n- Working on production codebases (>1000 lines)\n- Making changes that touch multiple files or components\n- The AI assistant starts \"suggesting improvements\" you didn't ask for\n- You notice the AI making assumptions about your intent\n\n## The 4 Foundational Rules (Karpathy)\n\n### Rule 1: Think Before Coding\n\n**Problem**: AI acts on assumptions, not understanding.\n\nBefore writing any code:\n1. If the requirement is ambiguous → **ask**, don't guess\n2. If there are multiple valid approaches → **list them** with tradeoffs\n3. If the request seems wrong → **push back** with reasoning\n4. If you're uncertain about scope → **confirm** before touching files\n\n❌ Bad: \"I'll refactor this module while fixing the bug\"\n✅ Good: \"The bug is in line 42. I see the surrounding code could be cleaner, but you only asked me to fix the bug. Should I also refactor?\"\n\n### Rule 2: Simplicity First\n\n**Problem**: AI defaults to over-abstraction.\n\n- 10 lines > 100 lines for the same result\n- No feature creep — only build what was asked\n- No premature abstraction — don't add interfaces \"just in case\"\n- Litmus test: would a senior engineer say \"this is too complex\"? → rewrite\n\n❌ Bad: Adding a factory pattern, three interfaces, and a config layer for a simple utility function\n✅ Good: One function, clear name, no unnecessary indirection\n\n### Rule 3: Surgical Changes\n\n**Problem**: AI makes \"drive-by\" edits to code it wasn't asked to touch.\n\n- Fix the bug, **only** the bug\n- Don't reformat adjacent code\n- Don't update comments you weren't asked about\n- Don't change variable names in unrelated functions\n- Every changed line must trace back to the user's specific request\n\n❌ Bad: \"While fixing the auth bug, I also cleaned up the logging format and renamed some variables\"\n✅ Good: 3 lines changed, all in the auth function, all directly related to the bug\n\n### Rule 4: Goal-Driven Execution\n\n**Problem**: Vague instructions lead to vague results.\n\nInstead of telling the AI **how** to do something, give it a **success criterion**:\n\n❌ \"Fix the login bug\"\n✅ \"Write a test that reproduces the login timeout on slow networks, then make it pass\"\n\n❌ \"Improve the API\"\n✅ \"Response time for /api/users must be under 200ms for 1000 concurrent requests\"\n\nThe AI iterates better toward measurable goals than fuzzy directions.\n\n> 💡 **Why This Way**: LLMs are natural iterators. Given a clear target, they'll loop (generate → test → adjust) until they hit it. Given a vague goal, they'll generate once, declare victory, and move on.\n\n## Battle-Tested Additions (Beyond Karpathy)\n\n### A1: Three-Layer Consistency Check\n\nAfter any change, verify alignment across layers:\n\n**Layer 1 — Naming**: env vars, DB columns, API paths, config keys must match across all files\n**Layer 2 — Business**: design docs ↔ code ↔ UI ↔ API responses must tell the same story\n**Layer 3 — Database**: migrations ordered correctly, FK references valid, types match TS interfaces\n\nRun the relevant layer after each change. Run all three on major releases.\n\n### A2: Anti-Rationalization\n\nNever trust the AI's \"I think this looks correct.\" \n\n- \"I read the code\" ≠ verified → **run it**\n- \"It should work\" ≠ confirmed → **test it**\n- \"I wrote it, so it's right\" = rationalization → **verify independently**\n\n### A3: Verification Loop\n\nFor every change type, define a verification action:\n\n| Changed | Verify by |\n|---|---|\n| Code/script | Execute it |\n| Config | Restart + confirm effect |\n| Generated file | Check content (wc -l, grep, diff) |\n| API call | Check return value |\n| UI change | Visual diff before/after |\n\n### A4: Pre-Change Snapshot\n\nBefore modifying any file:\n1. Record current state (grep key content, or screenshot)\n2. Make the change\n3. Diff to confirm only intended parts changed\n4. If unintended changes found → revert and redo surgically\n\n### A5: Context Hygiene\n\nAI context windows are finite. Polluted context → degraded output.\n\n- Trim tool outputs (pipe to `head -30`, don't dump 500 lines)\n- Checkpoint progress to files during long tasks\n- Don't let the AI \"remember\" — make it **read files**\n\n## Integration\n\n### Claude Code (CLAUDE.md)\nAdd to your project's `CLAUDE.md`:\n```\n# Engineering Discipline Rules\n[paste the 4 rules + additions above]\n```\n\n### Cursor (.cursor/rules)\nAdd to `.cursor/rules/engineering-discipline.md`\n\n### Any AI Coding Tool\nThese rules work as system prompts, project instructions, or conversation primers for any LLM-based coding assistant.\n\n## Related Skills\n- `trinity-harness` — Full agent harness with Challenge + Execute + Compound layers\n- `self-improving-agent` — Continuous learning from mistakes\n- `skill-creator` — Create new skills from workflows\n\n## Feedback\n- If useful: `clawhub star engineering-discipline`\n- Issues: https://github.com/clawhub/engineering-discipline\n\nFile v1.0.1:_meta.json\n\n{\n  \"ownerId\": \"kn757z7y3qhq5501ytv82v5p6d84ep08\",\n  \"slug\": \"engineering-discipline\",\n  \"version\": \"1.0.1\",\n  \"publishedAt\": 1781106955173\n}\n\nFile v1.0.1:skill-card.md\n\n## Description:\n\nEngineering Discipline provides coding-process guardrails that prompt an AI coding assistant to clarify ambiguous requests, keep changes scoped, avoid over-engineering, and verify work before completion.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[christianye](https://clawhub.ai/user/christianye)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nDevelopers and engineering teams use this skill to shape AI coding assistants during implementation, refactoring, and production code maintenance so outputs stay focused, minimal, and verified.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: Broad coding-process guardrails may add friction by increasing clarification, scope control, and verification steps during coding sessions.\n\nMitigation: Use this skill when that disciplined coding posture is desired, and review whether its clarification and verification expectations fit the project workflow before deployment.\n\nRisk: The guidance can limit unsolicited refactoring or broad edits, which may be inconvenient when exploratory redesign is the intended task.\n\nMitigation: State the intended refactor or exploration scope explicitly so the assistant can apply the guardrails without blocking the requested work.\n\n## Reference(s):\n\n- [ClawHub skill page](https://clawhub.ai/christianye/skills/engineering-discipline)\n- [Publisher profile](https://clawhub.ai/user/christianye)\n\n## Skill Output:\n\n**Output Type(s):** [Guidance, Markdown, Configuration]\n\n**Output Format:** [Markdown guidance with examples and configuration snippets]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [No hidden code, data access, or automatic persistence is disclosed in the security evidence.]\n\n## Skill Version(s):\n\n1.0.1 (source: frontmatter and 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 v1.0.0: 3 files, 4425 bytes\n\nFiles: skill-card.md (2252b), SKILL.md (5640b), _meta.json (141b)\n\nFile v1.0.0:SKILL.md\n\n---\nname: Engineering Discipline\nslug: engineering-discipline\nversion: 1.0.0\ndescription: \"Production-grade AI coding discipline. Prevents the top 4 AI coding sins: acting without thinking, over-engineering, collateral damage, and vague execution. Based on Karpathy's 4 rules + battle-tested additions (3-layer consistency checks, anti-rationalization, verification loops, surgical diffs). Works with Claude Code, Cursor, Copilot, and any AI coding assistant.\"\nhomepage: https://clawhub.ai/skills/engineering-discipline\nmetadata: {\"clawdbot\":{\"emoji\":\"🔧\",\"os\":[\"linux\",\"darwin\",\"win32\"]}}\n---\n\n## When to Use\n\nApply this skill to **every coding session**. It's not a task-specific tool — it's a permanent behavior modifier for AI coding assistants.\n\nEspecially critical when:\n- Working on production codebases (>1000 lines)\n- Making changes that touch multiple files or components\n- The AI assistant starts \"suggesting improvements\" you didn't ask for\n- You notice the AI making assumptions about your intent\n\n## The 4 Foundational Rules (Karpathy)\n\n### Rule 1: Think Before Coding\n\n**Problem**: AI acts on assumptions, not understanding.\n\nBefore writing any code:\n1. If the requirement is ambiguous → **ask**, don't guess\n2. If there are multiple valid approaches → **list them** with tradeoffs\n3. If the request seems wrong → **push back** with reasoning\n4. If you're uncertain about scope → **confirm** before touching files\n\n❌ Bad: \"I'll refactor this module while fixing the bug\"\n✅ Good: \"The bug is in line 42. I see the surrounding code could be cleaner, but you only asked me to fix the bug. Should I also refactor?\"\n\n### Rule 2: Simplicity First\n\n**Problem**: AI defaults to over-abstraction.\n\n- 10 lines > 100 lines for the same result\n- No feature creep — only build what was asked\n- No premature abstraction — don't add interfaces \"just in case\"\n- Litmus test: would a senior engineer say \"this is too complex\"? → rewrite\n\n❌ Bad: Adding a factory pattern, three interfaces, and a config layer for a simple utility function\n✅ Good: One function, clear name, no unnecessary indirection\n\n### Rule 3: Surgical Changes\n\n**Problem**: AI makes \"drive-by\" edits to code it wasn't asked to touch.\n\n- Fix the bug, **only** the bug\n- Don't reformat adjacent code\n- Don't update comments you weren't asked about\n- Don't change variable names in unrelated functions\n- Every changed line must trace back to the user's specific request\n\n❌ Bad: \"While fixing the auth bug, I also cleaned up the logging format and renamed some variables\"\n✅ Good: 3 lines changed, all in the auth function, all directly related to the bug\n\n### Rule 4: Goal-Driven Execution\n\n**Problem**: Vague instructions lead to vague results.\n\nInstead of telling the AI **how** to do something, give it a **success criterion**:\n\n❌ \"Fix the login bug\"\n✅ \"Write a test that reproduces the login timeout on slow networks, then make it pass\"\n\n❌ \"Improve the API\"\n✅ \"Response time for /api/users must be under 200ms for 1000 concurrent requests\"\n\nThe AI iterates better toward measurable goals than fuzzy directions.\n\n> 💡 **Why This Way**: LLMs are natural iterators. Given a clear target, they'll loop (generate → test → adjust) until they hit it. Given a vague goal, they'll generate once, declare victory, and move on.\n\n## Battle-Tested Additions (Beyond Karpathy)\n\n### A1: Three-Layer Consistency Check\n\nAfter any change, verify alignment across layers:\n\n**Layer 1 — Naming**: env vars, DB columns, API paths, config keys must match across all files\n**Layer 2 — Business**: design docs ↔ code ↔ UI ↔ API responses must tell the same story\n**Layer 3 — Database**: migrations ordered correctly, FK references valid, types match TS interfaces\n\nRun the relevant layer after each change. Run all three on major releases.\n\n### A2: Anti-Rationalization\n\nNever trust the AI's \"I think this looks correct.\" \n\n- \"I read the code\" ≠ verified → **run it**\n- \"It should work\" ≠ confirmed → **test it**\n- \"I wrote it, so it's right\" = rationalization → **verify independently**\n\n### A3: Verification Loop\n\nFor every change type, define a verification action:\n\n| Changed | Verify by |\n|---|---|\n| Code/script | Execute it |\n| Config | Restart + confirm effect |\n| Generated file | Check content (wc -l, grep, diff) |\n| API call | Check return value |\n| UI change | Visual diff before/after |\n\n### A4: Pre-Change Snapshot\n\nBefore modifying any file:\n1. Record current state (grep key content, or screenshot)\n2. Make the change\n3. Diff to confirm only intended parts changed\n4. If unintended changes found → revert and redo surgically\n\n### A5: Context Hygiene\n\nAI context windows are finite. Polluted context → degraded output.\n\n- Trim tool outputs (pipe to `head -30`, don't dump 500 lines)\n- Checkpoint progress to files during long tasks\n- Don't let the AI \"remember\" — make it **read files**\n\n## Integration\n\n### Claude Code (CLAUDE.md)\nAdd to your project's `CLAUDE.md`:\n```\n# Engineering Discipline Rules\n[paste the 4 rules + additions above]\n```\n\n### Cursor (.cursor/rules)\nAdd to `.cursor/rules/engineering-discipline.md`\n\n### Any AI Coding Tool\nThese rules work as system prompts, project instructions, or conversation primers for any LLM-based coding assistant.\n\n## Related Skills\n- `trinity-harness` — Full agent harness with Challenge + Execute + Compound layers\n- `self-improving-agent` — Continuous learning from mistakes\n- `skill-creator` — Create new skills from workflows\n\n## Feedback\n- If useful: `clawhub star engineering-discipline`\n- Issues: https://github.com/clawhub/engineering-discipline\n\nFile v1.0.0:_meta.json\n\n{\n  \"ownerId\": \"kn757z7y3qhq5501ytv82v5p6d84ep08\",\n  \"slug\": \"engineering-discipline\",\n  \"version\": \"1.0.0\",\n  \"publishedAt\": 1777023654273\n}\n\nFile v1.0.0:skill-card.md\n\n## Description: <br>\nProduction-grade AI coding discipline that gives AI coding assistants rules for thinking before coding, keeping changes simple and surgical, and verifying outcomes. <br>\n\nThis skill is ready for commercial/non-commercial use. <br>\n\n## Publisher: <br>\n[christianye](https://clawhub.ai/user/christianye) <br>\n\n### License/Terms of Use: <br>\nMIT-0 <br>\n\n\n## Use Case: <br>\nDevelopers and engineering teams use this skill as a standing behavior layer for AI coding sessions, especially when working on production codebases or changes that span files. It helps agents ask clarifying questions, avoid unnecessary refactors, keep diffs scoped, and verify changes with concrete checks. <br>\n\n### Deployment Geography for Use: <br>\nGlobal <br>\n\n## Known Risks and Mitigations: <br>\nRisk: Broad coding-session guidance can influence agent behavior across unrelated tasks if installed globally. <br>\nMitigation: Review the skill text before installation and use it only in coding contexts where disciplined scoping, verification, and clarification behavior is desired. <br>\nRisk: The security evidence reports no concrete findings, but notes that the VirusTotal check is still pending. <br>\nMitigation: Treat the clean scan as provisional until pending external checks complete, and avoid granting extra file, tool, or account access beyond the task. <br>\n\n\n## Reference(s): <br>\n- [ClawHub skill page](https://clawhub.ai/christianye/engineering-discipline) <br>\n- [Publisher profile](https://clawhub.ai/user/christianye) <br>\n\n\n## Skill Output: <br>\n**Output Type(s):** [guidance, text, markdown, shell commands] <br>\n**Output Format:** [Markdown guidance with occasional inline command examples] <br>\n**Output Parameters:** [1D] <br>\n**Other Properties Related to Output:** [The skill produces behavioral guidance for an agent rather than generated files or tool integrations.] <br>\n\n## Skill Version(s): <br>\n1.0.0 (source: frontmatter and release evidence) <br>\n\n## Ethical Considerations: <br>\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. <br>","readmeExcerpt":"Skill: Engineering Discipline Owner: christianye Summary: Production-grade AI coding discipline. Prevents the top 4 AI coding sins: acting without thinking, over-engineering, collateral damage, and vague execution.... Tags: latest:1.0.1 Version history: v1.0.1 | 2026-06-10T15:55:55.173Z | user Expand description with Chinese trigger keywords for better discoverability (反合理化 / 三层一致性检查 / Karpathy 四规则 / 开发纪律 etc.). No c","codeSnippets":[],"executableExamples":[{"language":"text","snippet":"# Engineering Discipline Rules\n[paste the 4 rules + additions above]"},{"language":"text","snippet":"# Engineering Discipline Rules\n[paste the 4 rules + additions above]"}],"parameters":null,"dependencies":[],"permissions":[],"extractedFiles":[{"path":"SKILL.md","content":"---\nname: Engineering Discipline\nslug: engineering-discipline\nversion: 1.0.1\ndescription: \"Production-grade AI coding discipline. Prevents the top 4 AI coding sins: acting without thinking, over-engineering, collateral damage, and vague execution. Triggers: 'engineering discipline', '过度工程', '反合理化', '三层一致性检查', 'AI coding rules', 'Karpathy 四规则', '写代码前先想', '手术刀修复'. Use when starting any coding task, before large refactors, or when catching yourself rationalizing shortcuts. Based on Karpathy's 4 rules + battle-tested additions (3-layer consistency checks, anti-rationalization, verification loops, surgical diffs). Works with Claude Code, Cursor, Copilot, OpenClaw, and any AI coding assistant. Triggers: 'engineering checklist'、'coding discipline'、'production quality'、'AI coding guardrails'、'开发纪律'。\"\nhomepage: https://clawhub.ai/skills/engineering-discipline\nmetadata: {\"clawdbot\":{\"emoji\":\"🔧\",\"os\":[\"linux\",\"darwin\",\"win32\"]}}\n---\n\n## When to Use\n\nApply this skill to **every coding session**. It's not a task-specific tool — it's a permanent behavior modifier for AI coding assistants.\n\nEspecially critical when:\n- Working on production codebases (>1000 lines)\n- Making changes that touch multiple files or components\n- The AI assistant starts \"suggesting improvements\" you didn't ask for\n- You notice the AI making assumptions about your intent\n\n## The 4 Foundational Rules (Karpathy)\n\n### Rule 1: Think Before Coding\n\n**Problem**: AI acts on assumptions, not understanding.\n\nBefore writing any code:\n1. If the requirement is ambiguous → **ask**, don't guess\n2. If there are multiple valid approaches → **list them** with tradeoffs\n3. If the request seems wrong → **push back** with reasoning\n4. If you're uncertain about scope → **confirm** before touching files\n\n❌ Bad: \"I'll refactor this module while fixing the bug\"\n✅ Good: \"The bug is in line 42. I see the surrounding code could be cleaner, but you only asked me to fix the bug. Should I also refactor?\"\n\n### Rule 2: Simplicity First\n\n**Problem**: AI defaults to over-abstraction.\n\n- 10 lines > 100 lines for the same result\n- No feature creep — only build what was asked\n- No premature abstraction — don't add interfaces \"just in case\"\n- Litmus test: would a senior engineer say \"this is too complex\"? → rewrite\n\n❌ Bad: Adding a factory pattern, three interfaces, and a config layer for a simple utility function\n✅ Good: One function, clear name, no unnecessary indirection\n\n### Rule 3: Surgical Changes\n\n**Problem**: AI makes \"drive-by\" edits to code it wasn't asked to touch.\n\n- Fix the bug, **only** the bug\n- Don't reformat adjacent code\n- Don't update comments you weren't asked about\n- Don't change variable names in unrelated functions\n- Every changed line must trace back to the user's specific request\n\n❌ Bad: \"While fixing the auth bug, I also cleaned up the logging format and renamed some variables\"\n✅ Good: 3 lines changed, all in the auth function, all directly related to the bug\n\n### Rule 4: Goal-Driven Execution\n\n**P"},{"path":"_meta.json","content":"{\n  \"ownerId\": \"kn757z7y3qhq5501ytv82v5p6d84ep08\",\n  \"slug\": \"engineering-discipline\",\n  \"version\": \"1.0.1\",\n  \"publishedAt\": 1781106955173\n}"},{"path":"skill-card.md","content":"## Description:\n\nEngineering Discipline provides coding-process guardrails that prompt an AI coding assistant to clarify ambiguous requests, keep changes scoped, avoid over-engineering, and verify work before completion.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[christianye](https://clawhub.ai/user/christianye)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nDevelopers and engineering teams use this skill to shape AI coding assistants during implementation, refactoring, and production code maintenance so outputs stay focused, minimal, and verified.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: Broad coding-process guardrails may add friction by increasing clarification, scope control, and verification steps during coding sessions.\n\nMitigation: Use this skill when that disciplined coding posture is desired, and review whether its clarification and verification expectations fit the project workflow before deployment.\n\nRisk: The guidance can limit unsolicited refactoring or broad edits, which may be inconvenient when exploratory redesign is the intended task.\n\nMitigation: State the intended refactor or exploration scope explicitly so the assistant can apply the guardrails without blocking the requested work.\n\n## Reference(s):\n\n- [ClawHub skill page](https://clawhub.ai/christianye/skills/engineering-discipline)\n- [Publisher profile](https://clawhub.ai/user/christianye)\n\n## Skill Output:\n\n**Output Type(s):** [Guidance, Markdown, Configuration]\n\n**Output Format:** [Markdown guidance with examples and configuration snippets]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [No hidden code, data access, or automatic persistence is disclosed in the security evidence.]\n\n## Skill Version(s):\n\n1.0.1 (source: frontmatter and 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."}],"languages":[],"docsSourceLabel":"CLAWHUB","editorialOverview":"Production-grade AI coding discipline. Prevents the top 4 AI coding sins: acting without thinking, over-engineering, collateral damage, and vague execution.... Skill: Engineering Discipline Owner: christianye Summary: Production-grade AI coding discipline. Prevents the top 4 AI coding sins: acting without thinking, over-engineering, collateral damage, and vague execution.... Tags: latest:1.0.1 Version history: v1.0.1 | 2026-06-10T15:55:55.173Z | user Expand description with Chinese trigger keywords for better discoverability (反合理化 / 三层一致性检查 / Karpathy 四规则 / 开发纪律 etc.). No c","editorialQuality":{"score":100,"threshold":65,"status":"ready","wordCount":1150,"uniquenessScore":55,"reasons":[]}},"media":{"evidence":{"source":"no-media","verified":false,"confidence":"low","updatedAt":"2026-10-11T17:29:07.734Z","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":"2026-10-11T17:29:07.734Z","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":"2026-10-11T20:58:03.493Z","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"}]}}}