bestdeejay-design/agent-skills / skills/docs-project
docs-project
UnverifiedProject (engineering) documentation branch — the 'how': README, ENTRY, AGENT/DELIVERY runbooks, and docs/ (REFERENCE map, ARCHITECTURE, ADR, SAGA, TEST_CASES, DEV_GUIDE, IMPROVEMENTS, TROUBLESHOOTING, BACKLOG, REVIEW, STATUS) plus contracts/ (OpenAPI/AsyncAPI — machine truth written BEFORE code). The map rule: every doc gets a card in REFERENCE.md. Full catalog + checklist in references/project-docs.md. Triggers: 'проектная документация', 'архитектура документация', 'adr', 'engineering docs', 'документация разработчика', 'контракты openapi', 'project docs', 'инженерная документация'.
Docs Project — engineering documentation («how»)
Use this skill to write the engineering documents: architecture, decisions
(ADR), contracts, dev guides, tests, and the documentation map. Assumes the
product docs exist (docs-product: VISION/PRD/ROADMAP/FEATURES).
When to use
- A project needs its engineering docs: README, ENTRY, ARCHITECTURE, ADR, tests.
- User asks for "проектная документация", "архитектура документация", "adr", "инженерная документация", "контракты openapi", "engineering docs".
- Contracts (OpenAPI/AsyncAPI) must be written before code.
Do NOT use
- For product docs (VISION/PRD/ROADMAP/FEATURES) — that is
docs-product. - For the meta-guide (phases, levels L1/L2/L3, completeness checklist) —
docs-system.
Root documents
| File | Purpose | Phase | Template |
|---|---|---|---|
README.md | Overview: what it is, stack, status, quick start, structure | 1 → 10 | templates/project/README.tmpl |
ENTRY.md | Entry point: "start here", navigation "topic → file", onboarding checklist | 9 | templates/project/ENTRY.tmpl |
AGENT.md | Runbook for the autonomous build agent: phases, rules, DoD, commit gate | 8 | — |
DELIVERY.md | Delivery gate: build/run/verify, what's in, what's NOT | 8 | — |
| Legal/community | LICENSE, CONTRIBUTING, CODE_OF_CONDUCT, SECURITY | 1/9 | — |
docs/ — engineering documents
| File | Purpose | Phase | Template |
|---|---|---|---|
docs/REFERENCE.md | The map. Card for EVERY doc: purpose/structure/facts/links, hierarchy of truth, drift table. Written last. | 10 | templates/project/REFERENCE.tmpl |
docs/ARCHITECTURE.md | Components, communication, data flows, tech stack | 2 → 6 | templates/project/ARCHITECTURE.tmpl |
docs/ADR/ADR-000.md | Architecture Decision Records — one per significant decision | 2 (ongoing) | templates/project/ADR-000.tmpl |
docs/SAGA.md | Cross-module scenarios (choreography, events, compensation, idempotency) | 5 | templates/project/SAGA.tmpl |
docs/TEST_CASES.md | Test cases (Gherkin-style) + coverage matrix | 6 | templates/project/TEST_CASES.tmpl |
docs/DEV_GUIDE.md | Local dev: prerequisites, run, env, migrations, debugging | 7 | templates/project/DEV_GUIDE.tmpl |
docs/IMPROVEMENTS.md | Known issues / doc-vs-fact drift + prioritized fix plan | 8 | templates/project/IMPROVEMENTS.tmpl |
docs/TROUBLESHOOTING.md | Runtime error diagnostics (E1…En) + startup checklist | 8 | templates/project/TROUBLESHOOTING.tmpl |
docs/BACKLOG.md | Ideas, deferred features, priorities | 1 | templates/project/BACKLOG.tmpl |
docs/REVIEW.md | Doc audit: problems, resolutions, per-service status | 10 | templates/project/REVIEW.tmpl |
docs/STATUS.md | L1 alternative to REVIEW: done / in progress / limitations | 10 | templates/project/STATUS.tmpl |
contracts/ — machine truth (API-first / microservices)
| Path | Purpose | When |
|---|---|---|
contracts/openapi/*.yaml | Per-service OpenAPI specs — the HTTP truth | before code |
contracts/asyncapi/events.yaml | Event catalog — the event truth | before code |
contracts/test/ | Contract-conformance test fixtures | with tests |
Contracts are written BEFORE code. The code implements the contract, not the other way around — «machine truth before code».
The map rule
Every document that exists gets a card in docs/REFERENCE.md
(purpose → structure → key facts → links). Written last, maintained on every
change: touch a doc → update its card.
Checklist (project side)
- One entry point exists (
ENTRY.md) and routes "topic → file". - Every doc has a card in
REFERENCE.md(including REFERENCE itself). - Hierarchy of truth declared (which doc wins on conflict).
- Drift table (doc-vs-fact) populated — no silent inconsistencies.
- Contracts exist before code for every API/event (API-first projects).
-
DELIVERY.mdstates how to build/verify/what's in/what's NOT.
Rule: engineering docs answer «how»; the map (
REFERENCE.md) is written last because it describes everything above it.
Boundaries
- Do not use for product vision, PRDs, or roadmap decisions; use
docs-product.
Authoring procedure
- Inventory the repository and establish the hierarchy of truth.
- Write contracts before implementation docs when the project exposes APIs or events.
- Draft architecture and ADRs from observed code and explicit decisions; label assumptions.
- Add runbooks and test cases with copy-pasteable commands and expected results.
- Write
REFERENCE.mdlast and add a card for every document.
Definition of done
Every generated document names its source of truth, owner or update trigger, prerequisites, commands, and failure path. Verify links, commands, and contract examples against the repository; do not mark a document complete merely because the template is filled.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-changelog-updatemorluto/reaPrepare or rewrite REA release changelogs and GitHub release notes from pinned Git history, with verified contributor thanks and Release Please synchronization. | Stars: 467k | Updated | Claude CodeCodex | Verified | Install |
| academy-guideanthropics/skillsStop and check this skill before finishing any reply to a question about how to use Claude or a Claude product — it recommends matching courses, tutorials, and use cases from Claude Academy (academy.claude.com), Anthropic's learning hub. Trigger on: "how do I", "how can I", "getting started with", "what can Claude do", "teach me", "learn to use"; questions about artifacts, projects, skills, plugins, connectors, MCP; requests about rolling Claude out to a team, class, or organization; and any ask for training materials, onboarding content, or learning resources. Use it when the user is learning how to use a feature or product — not when they are mid-task and just want the task done. This skill composes with other skills: after consulting product documentation to answer how a Claude feature works, also check here for a matching course or tutorial — a docs-grounded answer and an Academy recommendation belong together. Only recommend on a strong match; never invent Academy content. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| claude-apianthropics/skillsReference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent/MCP/tool-definition/multi-agent/RAG/LLM-judge/computer-use; generate/summarize/extract/classify/rewrite/converse over NL; debugging refusals/cutoffs/streaming/tool-calls/tokens). SKIP only when another provider is being worked on (overrides all triggers): OpenAI/GPT/Gemini/Llama/Mistral/Cohere/Ollama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST if no provider named — don't Read the file). | Stars: 196k | Updated | Claude CodeCodex | Unverified | Install |
| doc-coauthoringanthropics/skillsGuide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |