mgiovani/ai-arsenal / skills/docs-update
docs-update
VerifiedRefresh existing docs (architecture, onboarding, data-model, deployment, security, contributing) so they match the current codebase, verifying every claim against real code instead of guessing. Use when the user says docs are stale, asks to sync docs with recent code changes, update a specific doc file, or update a whole category (core/data/infrastructure/development) after a refactor or schema change. Not for creating docs that don't exist yet (use docs-init) or scoring/auditing doc health without editing (use docs-check).
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.8 KB |
| evals/evals.json | config | 2.7 KB |
| evals/trigger-eval.json | config | 1.9 KB |
| references/change-detection.md | markdown | 1.6 KB |
| references/update-strategies.md | markdown | 2.3 KB |
evals/evals.json
View on GitHub{
"skill": "docs-update",
"evals": [
{
"id": "architecture-update-after-refactor",
"prompt": "docs-update architecture — we just split the monolith into 3 services (auth, billing, notifications), can you sync docs/architecture.md?",
"assertions": [
"Reads the full existing docs/architecture.md and greps the codebase to verify the claimed service names/counts before writing, rather than trusting the old doc text",
"Runs a freshness check using docs/architecture.md's own git history (the resolved path from the argument), not a different or hardcoded document",
"Before filling any {{PLACEHOLDER}} tokens, determines which tokens the architecture template actually contains (e.g. via grep) rather than assuming a fixed list",
"Preserves any custom/manually-written sections in docs/architecture.md that aren't part of the template structure",
"Final report states the verified service count/names as facts traceable to a command run this session, not vague language like 'several services'"
]
},
{
"id": "multi-doc-update-with-unverifiable-claim",
"prompt": "docs-update all — sync everything with the current codebase, we've made a bunch of changes over the last month",
"assertions": [
"Resolves scope to every doc currently present in docs/ and processes each one (concurrently via one subagent per doc if a Task tool is available, otherwise sequentially)",
"For at least one document, verifies claims against the codebase via grep/find rather than assuming the existing text is correct",
"When a specific claim in a document cannot be confirmed in the codebase (e.g. an SSO/auth claim with no matching code), removes or flags it instead of leaving it unverified or inventing supporting detail",
"Final report distinguishes updated docs from up-to-date/skipped docs, and does not claim a doc was 'updated' if nothing in it actually changed",
"Does not overwrite sections identified as custom/manual content"
]
},
{
"id": "update-request-for-nonexistent-doc",
"prompt": "docs-update security — update our security documentation (docs/security.md does not exist in this repo)",
"assertions": [
"Recognizes docs/security.md does not currently exist rather than silently treating the request as a from-scratch creation",
"Does not write or create a new docs/security.md file with fabricated content",
"No mutating Write/Edit command is run against docs/security.md during this run",
"Final message tells the user to use docs-init (or equivalent) to create the document first, and stops there rather than generating one anyway"
]
}
]
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |