[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C0E9BXMMP6Q0B8EW61DXEJ":3,"skill-same:01M7C0E9BXMMP6Q0B8EW61DXEJ":44},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":26,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M7C0E9BXMMP6Q0B8EW61DXEJ","validation-methodology",null,"Cross-cutting validation discipline. Acceptance criteria format, definition of done checklists, quality gate criteria, verification methods. Loaded by every analysis agent.","zak-keown\u002Fmoe","zak-keown","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F320717241?v=4","01M7BZFXME0NW6TQX0H7AZM2KS","01M7BZFXN32VR0JN5AATRD53H4",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe-backstory\u002F.kimi-plugin\u002Fskills\u002Fvalidation-methodology\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: validation-methodology\ndescription: Cross-cutting validation discipline. Acceptance criteria format, definition of done checklists, quality gate criteria, verification methods. Loaded by every analysis agent.\n---\n\n# Validation Methodology\n\nSpecifications are claims. Claims require proof. This skill defines how the pipeline transforms claims into proof.\n\n## Acceptance Criteria Format\n\nEvery module spec receives formal acceptance criteria during Layer 4. Criteria use Given\u002FWhen\u002FThen format:\n\n```markdown\n### AC-CONFIG-001: Config File Load\n**Given** a valid config file exists at the documented path\n**When** the application starts\n**Then**\n- The application reads every declared key from the file\n- Unknown keys are logged as warnings, not errors\n- Missing optional keys fall back to documented defaults\n\n**Source:** official-docs, source-code, runtime-observation\n**Confidence:** confirmed\n**Priority:** P0\n**Verification:** Automated test\n```\n\n### ID Format: `AC-{DOMAIN}-{NNN}`\n\n- `{DOMAIN}`: Short uppercase token for the behavioral domain (matches the spec's domain tokens — e.g., `CLI`, `CONFIG`, `API`, `AUDIO`, `PARSER`)\n- `{NNN}`: Zero-padded three-digit sequence (001, 002, ...)\n- IDs are immutable — retired, never reused\n\n### Field Rules\n\n**Given\u002FWhen\u002FThen:**\n- Use observable terms, not implementation details\n- \"Given a valid config file exists\" not \"Given the ConfigLoader is initialized\"\n- Each Then outcome must be verifiable without implementation internals\n- Describe observable state, not UI mechanisms — \"system rejects malformed input\" not \"error banner is visible.\" Criteria must be testable across any frontend platform.\n\n**Priority:**\n- P0: Must pass. Blocks pipeline. Implementation cannot ship without satisfying.\n- P1: Should pass. Reported as warnings. Divergence requires documented rationale.\n- P2: Nice to pass. Advisory.\n\n**Verification:** One of the 5 methods below.\n\n### Output Location\n\n```\nworkspace\u002Fraw\u002Fspecs\u002Fvalidation\u002Facceptance-criteria\u002F\n    _index.md                   # Summary with counts and links\n    config-loading.md           # AC-CONFIG-001 through AC-CONFIG-NNN\n    cli-flags.md                # AC-CLI-001 through ...\n    ...\n```\n\n## Verification Methods\n\nEvery acceptance criterion specifies how the implementer will verify it:\n\n| Method | ID | When to Use |\n|--------|-----|------------|\n| **Automated Test** | `automated-test` | Precondition, action, and outcome are all programmable |\n| **Manual Inspection** | `manual-inspection` | Outcome requires human judgment (UX quality, clarity) |\n| **Runtime Observation** | `runtime-observation` | Behavior involves timing, async events, sustained observation |\n| **Protocol Capture** | `protocol-capture` | Wire format, headers, API protocol compliance |\n| **Code Review** | `code-review` | Architectural or quality criteria not directly testable at runtime |\n\n## Quality Gate Criteria\n\n```dot\ndigraph gate_evaluation {\n    rankdir=TB;\n\n    \"Agent produces output\" [shape=ellipse];\n    \"Check DOD checklist for agent layer\" [shape=box];\n    \"All DOD items pass?\" [shape=diamond];\n    \"Proceed to gate evaluation\" [shape=box];\n    \"Flag incomplete output for review\" [shape=box];\n    \"Check gate criteria\" [shape=box];\n    \"All gate criteria pass?\" [shape=diamond];\n    \"Tag and proceed to next layer\" [shape=box];\n    \"Remediate findings\" [shape=box];\n    \"Attempts exhausted?\" [shape=diamond];\n    \"STOP: Pipeline blocked\" [shape=octagon, style=filled, fillcolor=red, fontcolor=white];\n\n    \"Agent produces output\" -> \"Check DOD checklist for agent layer\";\n    \"Check DOD checklist for agent layer\" -> \"All DOD items pass?\";\n    \"All DOD items pass?\" -> \"Proceed to gate evaluation\" [label=\"yes\"];\n    \"All DOD items pass?\" -> \"Flag incomplete output for review\" [label=\"no\"];\n    \"Flag incomplete output for review\" -> \"Proceed to gate evaluation\";\n    \"Proceed to gate evaluation\" -> \"Check gate criteria\";\n    \"Check gate criteria\" -> \"All gate criteria pass?\";\n    \"All gate criteria pass?\" -> \"Tag and proceed to next layer\" [label=\"yes\"];\n    \"All gate criteria pass?\" -> \"Remediate findings\" [label=\"no\"];\n    \"Remediate findings\" -> \"Attempts exhausted?\";\n    \"Attempts exhausted?\" -> \"STOP: Pipeline blocked\" [label=\"yes\"];\n    \"Attempts exhausted?\" -> \"Check gate criteria\" [label=\"no\"];\n}\n```\n\n### Gate 1 (spec-verifier, after Layer 3)\n\n- [ ] Zero contradictions between specs\n- [ ] All crypto claims and constants verified against source (not assumed)\n- [ ] Every behavioral claim has a provenance citation\n- [ ] Assumed claims are a small minority — most claims have direct evidence. A module dominated by assumptions needs more intelligence gathering before proceeding\n- [ ] All modules have specs\n- [ ] Zero uncited behavioral claims in output candidates\n\n### Gate 2 (spec-reviewer, after Layer 4)\n\n- [ ] Zero implementation leakage in specs\n- [ ] Zero P0 completeness gaps\n- [ ] All ACs have IDs and are linked to specs\n- [ ] P0 ACs have test vectors\n- [ ] All ACs are testable\n- [ ] All modules have ACs\n- [ ] Gap report reviewed\n- [ ] Zero contamination in output\n\n### Implementation Leakage Definition\n\n\"Zero implementation leakage\" means every identifier, name, and reference in the output specs passes the reimplementor test: **\"Could the reimplementor reasonably redesign this?\"** If yes, it's an implementation detail that should have been abstracted. If no, it's an external contract that belongs in the spec.\n\n**Implementation details (must not appear in output specs):**\n\n| Category | What to look for |\n|----------|-----------------|\n| Internal names | Variable names, function signatures, class names, method names from the source — any language |\n| Internal architecture | Module boundaries, file organization, inheritance hierarchies described in prose |\n| Framework-specific patterns | State management internals, ORM patterns, framework lifecycle hooks by internal name |\n| Build\u002Fdeployment artifacts | Source file paths, line numbers, chunk IDs, minified identifiers |\n| Code structure language | \"calls X then Y\", \"inherits from Z\", \"implements interface W\" |\n| Internal feature gates | Feature flags, A\u002FB test names, internal telemetry event names |\n\n**External contracts (must be preserved):**\n\n| Category | What to look for |\n|----------|-----------------|\n| User-facing identifiers | CLI flags, env vars, config keys, config file paths |\n| Wire protocol fields | API endpoints, request\u002Fresponse field names, header names |\n| External system schemas | Database tables\u002Fcolumns (shared), third-party API contracts, external CLI tool flags |\n| Published constants | Error message text, exit codes, timeout values |\n| Standard names | Protocol names (OAuth, JWT), encoding names (UTF-8), algorithm names (SHA-256) |\n\nSee the **spec-sanitization skill** for the full semantic classification (Implementation Detail vs. External Contract) and rewrite guidance.\n\n## Definition of Done Checklists\n\n### Layer 1 Agent DOD\n\n- [ ] All output files at correct workspace directory for source origin\n- [ ] All behavioral claims have `\u003C!-- cite: -->` provenance citations\n- [ ] Output follows standard format for this agent's intelligence source\n- [ ] Session JSONL captured to `workspace\u002Fprovenance\u002Fsessions\u002F`\n- [ ] No raw source code excerpts longer than one line\n\n### Layer 2 Agent DOD\n\n- [ ] Feature inventory covers all sources consulted\n- [ ] Architecture document covers all identified components\n- [ ] API surface covers all discovered interfaces\n- [ ] Module map is complete (name, description, priority, dependencies)\n- [ ] No known gaps (or gaps explicitly documented with `[GAP]` marker)\n- [ ] Cross-references to Layer 1 artifacts are valid\n- [ ] Session JSONL captured\n\n### Layer 3 Agent DOD\n\n- [ ] Spec covers: entry points, decision trees, state machines, error handling, edge cases\n- [ ] Zero implementation details (no function names, variable names, line numbers)\n- [ ] All behavioral claims have provenance citations\n- [ ] Assumed claims are a small minority; most claims have direct evidence\n- [ ] Self-assessment verification checklist completed at end of spec\n- [ ] Session JSONL captured\n\n### Layer 3 Spec Self-Assessment Checklist\n\nEvery Layer 3 spec file ends with:\n\n```markdown\n## Verification Checklist\n- [ ] All entry points documented\n- [ ] All decision trees traced (every branch, including error paths)\n- [ ] All state machines documented (states, transitions, triggers)\n- [ ] All error conditions listed with observable symptoms\n- [ ] All edge cases identified and documented\n- [ ] Zero source code identifiers in this document\n- [ ] All behavioral claims have provenance citations\n- [ ] Assumed claims are a small minority; most claims have direct evidence\n- [ ] No sections are empty or placeholder-only\n- [ ] Cross-references to other specs are valid\n```\n\n### Layer 4 Agent DOD\n\n- [ ] Test vectors have concrete input\u002Foutput pairs (not abstract descriptions)\n- [ ] Runnable test harness scripts produced where feasible\n- [ ] ACs have `AC-{MODULE}-{NNN}` IDs and Given\u002FWhen\u002FThen structure\n- [ ] Every AC has a verification method assigned\n- [ ] Every AC has provenance links\n- [ ] Every P0 module has test vectors and ACs\n- [ ] Session JSONL captured\n\n### Gate Agent DOD\n\n- [ ] Report written to correct path\n- [ ] Clear PASS or FAIL status in summary\n- [ ] Every failing criterion individually documented\n- [ ] Quantitative summary (claims checked, contradictions, gaps, coverage %)\n- [ ] Session JSONL captured\n\n### Layer 5 Agent DOD\n\n- [ ] All raw specs have clean counterparts in `workspace\u002Foutput\u002F`\n- [ ] Zero source code identifiers in the output files\n- [ ] All 17 contamination detection patterns return 0 on `workspace\u002Foutput\u002F` (see Implementation Leakage Definition)\n- [ ] All behavioral information preserved (feature count, AC count identical)\n- [ ] Provenance metadata preserved (confidence, source types present)\n- [ ] Validation artifacts sanitized and copied to the output\n- [ ] Session JSONL captured\n\n### Layer 6 Agent DOD (Second-Pass Review)\n\n- [ ] All 8 structural leakage checks executed\n- [ ] All 12 content contamination checks executed\n- [ ] All 9 behavioral completeness checks executed\n- [ ] Detailed audit reports written to `workspace\u002Fraw\u002Faudit\u002F`\n- [ ] Clean audit summaries written to `workspace\u002Foutput\u002Faudit\u002F` (outcomes only)\n- [ ] Overall PASS determination (or FAIL after 3 remediation attempts)\n- [ ] `review-complete` git tag applied\n- [ ] Session JSONL captured\n\n## Confidence Principles\n\nThe goal is specs you'd trust enough to implement from. Apply judgment, not arithmetic:\n\n- **Every claim needs provenance.** An uncited behavioral claim is an unverified claim. Zero uncited claims in output — this is a hard gate.\n- **Assumed claims are a warning sign.** A few assumptions are inevitable (indirect reasoning, naming conventions). But a module where most claims are assumed hasn't been analyzed — it's been guessed at. Gather more intelligence.\n- **Confirmed claims build trust.** Claims corroborated by multiple independent sources (source + runtime, source + docs) are the foundation. Critical modules should be dominated by confirmed claims.\n- **P0 behaviors need test vectors.** Every P0 acceptance criterion must have a concrete test vector. No exceptions.\n\n## When Confidence Is Low\n\n- **Too many assumptions:** Dispatch additional intelligence-gathering agents. Try a different mode (runtime observation often confirms what source analysis can only infer).\n- **Claims lack corroboration:** Check if other modes have evidence. A claim from source alone is `inferred`; the same claim confirmed by runtime observation becomes `confirmed`.\n- **Uncited claims in output\u002F:** Hard failure — pipeline blocked until resolved. Trace each uncited claim back to its originating agent and require a citation.\n","# Validation Methodology\n\nSpecifications are claims. Claims require proof. This skill defines how the pipeline transforms claims into proof.\n\n## Acceptance Criteria Format\n\nEvery module spec receives formal acceptance criteria during Layer 4. Criteria use Given\u002FWhen\u002FThen format:\n\n```markdown\n### AC-CONFIG-001: Config File Load\n**Given** a valid config file exists at the documented path\n**When** the application starts\n**Then**\n- The application reads every declared key from the file\n- Unknown keys are logged as warnings, not errors\n- Missing optional keys fall back to documented defaults\n\n**Source:** official-docs, source-code, runtime-observation\n**Confidence:** confirmed\n**Priority:** P0\n**Verification:** Automated test\n```\n\n### ID Format: `AC-{DOMAIN}-{NNN}`\n\n- `{DOMAIN}`: Short uppercase token for the behavioral domain (matches the spec's domain tokens — e.g., `CLI`, `CONFIG`, `API`, `AUDIO`, `PARSER`)\n- `{NNN}`: Zero-padded three-digit sequence (001, 002, ...)\n- IDs are immutable — retired, never reused\n\n### Field Rules\n\n**Given\u002FWhen\u002FThen:**\n- Use observable terms, not implementation details\n- \"Given a valid config file exists\" not \"Given the ConfigLoader is initialized\"\n- Each Then outcome must be verifiable without implementation internals\n- Describe observable state, not UI mechanisms — \"system rejects malformed input\" not \"error banner is visible.\" Criteria must be testable across any frontend platform.\n\n**Priority:**\n- P0: Must pass. Blocks pipeline. Implementation cannot ship without satisfying.\n- P1: Should pass. Reported as warnings. Divergence requires documented rationale.\n- P2: Nice to pass. Advisory.\n\n**Verification:** One of the 5 methods below.\n\n### Output Location\n\n```\nworkspace\u002Fraw\u002Fspecs\u002Fvalidation\u002Facceptance-criteria\u002F\n    _index.md                   # Summary with counts and links\n    config-loading.md           # AC-CONFIG-001 through AC-CONFIG-NNN\n    cli-flags.md                # AC-CLI-001 through ...\n    ...\n```\n\n## Verification Methods\n\nEvery acceptance criterion specifies how the implementer will verify it:\n\n| Method | ID | When to Use |\n|--------|-----|------------|\n| **Automated Test** | `automated-test` | Precondition, action, and outcome are all programmable |\n| **Manual Inspection** | `manual-inspection` | Outcome requires human judgment (UX quality, clarity) |\n| **Runtime Observation** | `runtime-observation` | Behavior involves timing, async events, sustained observation |\n| **Protocol Capture** | `protocol-capture` | Wire format, headers, API protocol compliance |\n| **Code Review** | `code-review` | Architectural or quality criteria not directly testable at runtime |\n\n## Quality Gate Criteria\n\n```dot\ndigraph gate_evaluation {\n    rankdir=TB;\n\n    \"Agent produces output\" [shape=ellipse];\n    \"Check DOD checklist for agent layer\" [shape=box];\n    \"All DOD items pass?\" [shape=diamond];\n    \"Proceed to gate evaluation\" [shape=box];\n    \"Flag incomplete output for review\" [shape=box];\n    \"Check gate criteria\" [shape=box];\n    \"All gate criteria pass?\" [shape=diamond];\n    \"Tag and proceed to next layer\" [shape=box];\n    \"Remediate findings\" [shape=box];\n    \"Attempts exhausted?\" [shape=diamond];\n    \"STOP: Pipeline blocked\" [shape=octagon, style=filled, fillcolor=red, fontcolor=white];\n\n    \"Agent produces output\" -> \"Check DOD checklist for agent layer\";\n    \"Check DOD checklist for agent layer\" -> \"All DOD items pass?\";\n    \"All DOD items pass?\" -> \"Proceed to gate evaluation\" [label=\"yes\"];\n    \"All DOD items pass?\" -> \"Flag incomplete output for review\" [label=\"no\"];\n    \"Flag incomplete output for review\" -> \"Proceed to gate evaluation\";\n    \"Proceed to gate evaluation\" -> \"Check gate criteria\";\n    \"Check gate criteria\" -> \"All gate criteria pass?\";\n    \"All gate criteria pass?\" -> \"Tag and proceed to next layer\" [label=\"yes\"];\n    \"All gate criteria pass?\" -> \"Remediate findings\" [label=\"no\"];\n    \"Remediate findings\" -> \"Attempts exhausted?\";\n    \"Attempts exhausted?\" -> \"STOP: Pipeline blocked\" [label=\"yes\"];\n    \"Attempts exhausted?\" -> \"Check gate criteria\" [label=\"no\"];\n}\n```\n\n### Gate 1 (spec-verifier, after Layer 3)\n\n- [ ] Zero contradictions between specs\n- [ ] All crypto claims and constants verified against source (not assumed)\n- [ ] Every behavioral claim has a provenance citation\n- [ ] Assumed claims are a small minority — most claims have direct evidence. A module dominated by assumptions needs more intelligence gathering before proceeding\n- [ ] All modules have specs\n- [ ] Zero uncited behavioral claims in output candidates\n\n### Gate 2 (spec-reviewer, after Layer 4)\n\n- [ ] Zero implementation leakage in specs\n- [ ] Zero P0 completeness gaps\n- [ ] All ACs have IDs and are linked to specs\n- [ ] P0 ACs have test vectors\n- [ ] All ACs are testable\n- [ ] All modules have ACs\n- [ ] Gap report reviewed\n- [ ] Zero contamination in output\n\n### Implementation Leakage Definition\n\n\"Zero implementation leakage\" means every identifier, name, and reference in the output specs passes the reimplementor test: **\"Could the reimplementor reasonably redesign this?\"** If yes, it's an implementation detail that should have been abstracted. If no, it's an external contract that belongs in the spec.\n\n**Implementation details (must not appear in output specs):**\n\n| Category | What to look for |\n|----------|-----------------|\n| Internal names | Variable names, function signatures, class names, method names from the source — any language |\n| Internal architecture | Module boundaries, file organization, inheritance hierarchies described in prose |\n| Framework-specific patterns | State management internals, ORM patterns, framework lifecycle hooks by internal name |\n| Build\u002Fdeployment artifacts | Source file paths, line numbers, chunk IDs, minified identifiers |\n| Code structure language | \"calls X then Y\", \"inherits from Z\", \"implements interface W\" |\n| Internal feature gates | Feature flags, A\u002FB test names, internal telemetry event names |\n\n**External contracts (must be preserved):**\n\n| Category | What to look for |\n|----------|-----------------|\n| User-facing identifiers | CLI flags, env vars, config keys, config file paths |\n| Wire protocol fields | API endpoints, request\u002Fresponse field names, header names |\n| External system schemas | Database tables\u002Fcolumns (shared), third-party API contracts, external CLI tool flags |\n| Published constants | Error message text, exit codes, timeout values |\n| Standard names | Protocol names (OAuth, JWT), encoding names (UTF-8), algorithm names (SHA-256) |\n\nSee the **spec-sanitization skill** for the full semantic classification (Implementation Detail vs. External Contract) and rewrite guidance.\n\n## Definition of Done Checklists\n\n### Layer 1 Agent DOD\n\n- [ ] All output files at correct workspace directory for source origin\n- [ ] All behavioral claims have `\u003C!-- cite: -->` provenance citations\n- [ ] Output follows standard format for this agent's intelligence source\n- [ ] Session JSONL captured to `workspace\u002Fprovenance\u002Fsessions\u002F`\n- [ ] No raw source code excerpts longer than one line\n\n### Layer 2 Agent DOD\n\n- [ ] Feature inventory covers all sources consulted\n- [ ] Architecture document covers all identified components\n- [ ] API surface covers all discovered interfaces\n- [ ] Module map is complete (name, description, priority, dependencies)\n- [ ] No known gaps (or gaps explicitly documented with `[GAP]` marker)\n- [ ] Cross-references to Layer 1 artifacts are valid\n- [ ] Session JSONL captured\n\n### Layer 3 Agent DOD\n\n- [ ] Spec covers: entry points, decision trees, state machines, error handling, edge cases\n- [ ] Zero implementation details (no function names, variable names, line numbers)\n- [ ] All behavioral claims have provenance citations\n- [ ] Assumed claims are a small minority; most claims have direct evidence\n- [ ] Self-assessment verification checklist completed at end of spec\n- [ ] Session JSONL captured\n\n### Layer 3 Spec Self-Assessment Checklist\n\nEvery Layer 3 spec file ends with:\n\n```markdown\n## Verification Checklist\n- [ ] All entry points documented\n- [ ] All decision trees traced (every branch, including error paths)\n- [ ] All state machines documented (states, transitions, triggers)\n- [ ] All error conditions listed with observable symptoms\n- [ ] All edge cases identified and documented\n- [ ] Zero source code identifiers in this document\n- [ ] All behavioral claims have provenance citations\n- [ ] Assumed claims are a small minority; most claims have direct evidence\n- [ ] No sections are empty or placeholder-only\n- [ ] Cross-references to other specs are valid\n```\n\n### Layer 4 Agent DOD\n\n- [ ] Test vectors have concrete input\u002Foutput pairs (not abstract descriptions)\n- [ ] Runnable test harness scripts produced where feasible\n- [ ] ACs have `AC-{MODULE}-{NNN}` IDs and Given\u002FWhen\u002FThen structure\n- [ ] Every AC has a verification method assigned\n- [ ] Every AC has provenance links\n- [ ] Every P0 module has test vectors and ACs\n- [ ] Session JSONL captured\n\n### Gate Agent DOD\n\n- [ ] Report written to correct path\n- [ ] Clear PASS or FAIL status in summary\n- [ ] Every failing criterion individually documented\n- [ ] Quantitative summary (claims checked, contradictions, gaps, coverage %)\n- [ ] Session JSONL captured\n\n### Layer 5 Agent DOD\n\n- [ ] All raw specs have clean counterparts in `workspace\u002Foutput\u002F`\n- [ ] Zero source code identifiers in the output files\n- [ ] All 17 contamination detection patterns return 0 on `workspace\u002Foutput\u002F` (see Implementation Leakage Definition)\n- [ ] All behavioral information preserved (feature count, AC count identical)\n- [ ] Provenance metadata preserved (confidence, source types present)\n- [ ] Validation artifacts sanitized and copied to the output\n- [ ] Session JSONL captured\n\n### Layer 6 Agent DOD (Second-Pass Review)\n\n- [ ] All 8 structural leakage checks executed\n- [ ] All 12 content contamination checks executed\n- [ ] All 9 behavioral completeness checks executed\n- [ ] Detailed audit reports written to `workspace\u002Fraw\u002Faudit\u002F`\n- [ ] Clean audit summaries written to `workspace\u002Foutput\u002Faudit\u002F` (outcomes only)\n- [ ] Overall PASS determination (or FAIL after 3 remediation attempts)\n- [ ] `review-complete` git tag applied\n- [ ] Session JSONL captured\n\n## Confidence Principles\n\nThe goal is specs you'd trust enough to implement from. Apply judgment, not arithmetic:\n\n- **Every claim needs provenance.** An uncited behavioral claim is an unverified claim. Zero uncited claims in output — this is a hard gate.\n- **Assumed claims are a warning sign.** A few assumptions are inevitable (indirect reasoning, naming conventions). But a module where most claims are assumed hasn't been analyzed — it's been guessed at. Gather more intelligence.\n- **Confirmed claims build trust.** Claims corroborated by multiple independent sources (source + runtime, source + docs) are the foundation. Critical modules should be dominated by confirmed claims.\n- **P0 behaviors need test vectors.** Every P0 acceptance criterion must have a concrete test vector. No exceptions.\n\n## When Confidence Is Low\n\n- **Too many assumptions:** Dispatch additional intelligence-gathering agents. Try a different mode (runtime observation often confirms what source analysis can only infer).\n- **Claims lack corroboration:** Check if other modes have evidence. A claim from source alone is `inferred`; the same claim confirmed by runtime observation becomes `confirmed`.\n- **Uncited claims in output\u002F:** Hard failure — pipeline blocked until resolved. Trace each uncited claim back to its originating agent and require a citation.\n",{"name":5,"description":7},"main","728894f3f85fbd4e788f015ddfd72bb6155d5dd5","plugins\u002Fmoe-backstory\u002F.kimi-plugin\u002Fskills\u002Fvalidation-methodology",1,"https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe","https:\u002F\u002Fgithub.com\u002Fzak-keown","2026-09-05 06:29:06",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add zak-keown\u002Fmoe --skill validation-methodology",{"tool":36,"command":37},"bunx","bunx skills add zak-keown\u002Fmoe --skill validation-methodology",{"tool":39,"command":40},"pnpm","pnpm dlx skills add zak-keown\u002Fmoe --skill validation-methodology","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe-backstory\u002F.kimi-plugin\u002Fskills\u002Fvalidation-methodology\u002FSKILL.md",true,[],[45,52,59,66,73,80],{"skillNo":46,"name":47,"nameZh":6,"description":48,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":49,"categories":50,"keywords":51,"publishedAt":18},"01M7C1Z5362Y8NVRZ69HWP0ZD0","write-skill","Use when creating new skills, editing existing skills, or verifying skills work before deployment","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwrite-skill\u002FSKILL.md",[],[],{"skillNo":53,"name":54,"nameZh":6,"description":55,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":56,"categories":57,"keywords":58,"publishedAt":18},"01M7C1YZ9966WK9R6ZRPCGWF3G","write-plan","Use when you have a spec or requirements for a multi-step task, before touching code","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwrite-plan\u002FSKILL.md",[],[],{"skillNo":60,"name":61,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":18},"01M7C1YMR4SJD7BPBBPXZCVGVH","write-clearly","Apply Strunk's timeless writing rules to ANY prose humans will read—documentation, commit messages, error messages, explanations, reports, or UI text. Makes your writing clearer, stronger, and more professional.","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwrite-clearly\u002FSKILL.md",[],[],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":18},"01M7C1YH9805YSM0N20VQAVSXF","windows-vm","Create, manage, or connect to a headless Windows 11 VM running in Docker with SSH access. Use when the user wants to spin up, stop, restart, or SSH into a Windows VM.","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwindows-vm\u002FSKILL.md",[],[],{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":18},"01M7C1YD87SCNEJ2C9B8HFR2NQ","verify-completion","Use when about to claim work is complete, fixed, or passing, before committing or creating MRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fverify-completion\u002FSKILL.md",[],[],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":18},"01M7C1YARDEHT0TJ9BRG30ZCF4","using-moe","Use when starting any conversation - establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fusing-moe\u002FSKILL.md",[],[]]