Skip to content

YurunChen/repo-docs-skills / skills/repo-docs

repo-docs

Verified

Build and maintain a Markdown guide that helps humans understand a repository through real behavior, concepts, and evidence. Use when a user asks to understand a repo, generate or update repo-docs, answer repo-architecture/onboarding questions, seed docs for a new project, sync docs after code changes, or delete generated repo docs.

FileKindSize
EXAMPLES.mdmarkdown33.1 KB
PAGE_RULES.mdmarkdown44.2 KB
QUALITY_RULES.mdmarkdown4.8 KB
REFERENCE.mdmarkdown1.0 KB
ROOT_AGENT_RULES.mdmarkdown4.2 KB
SCOPE_MODES.mdmarkdown4.9 KB
SKILL.mdskill md12.4 KB
SYNC_RULES.mdmarkdown20.5 KB
WRITING.mdmarkdown15.0 KB
evals/README.mdmarkdown3.9 KB
evals/fixtures/README.mdmarkdown294 B
evals/run_eval.pyscript20.3 KB
scripts/validate_repo_docs.pyexecutablescript46.6 KB
validate_repo_docs.pyscript482 B

SCOPE_MODES.md

View on GitHub
# Repo-Docs Scope And Modes

## Seed Project Mode

Use Seed Project Mode when the repo is new, empty, or only contains planning files. This mode exists so `repo-docs` can start maintaining project understanding before implementation exists.

### Detection

Treat a project as seed-stage when most of these are true:

- no substantive source files
- no runtime entrypoint, script, package manifest, or runnable command
- no tests, generated artifacts, schemas, or data contracts
- only README, license, `.gitignore`, root agent instructions, planning notes, or empty directories exist
- the user is asking to start, plan, initialize, or hand off a project before implementation

If the repo has a small amount of code but no stable workflow yet, use mixed mode: document implemented facts as `Confirmed`, planned behavior as `Planned`, and uncertain design as `Unknown` / `未确认`.

### Evidence and status labels

Seed docs use the same status family as the rest of the package (see Evidence Standard). They must separate intent from implementation:

| Label | Use when | Example wording |
| --- | --- | --- |
| `Confirmed` / `已确认` | Existing source, config, test, data, artifact, or command proves it. | `Evidence status: Confirmed. README.md exists and defines the project goal.` |
| `Planned` / `计划中` | A decided direction, proposed workflow, module, contract, or next step that is not implemented yet. | `Planned: the first version will target agent-compatible skills.` |
| `Unknown` / `未确认` | The project has not decided or verified it. | `Unknown: no test strategy has been chosen yet.` |

Do not convert chat intent into current-state facts. A user saying "we should add a CLI" is `Planned` unless a CLI exists. When a direction is explicitly locked in, mark it `Planned` and say "decided:" in prose; keep `Planned` until code proves it.

### Seed guide shape

For an empty project, use this status-labeled pack:

```text
repo-docs/
  README.md
  change-log.md
  glossary.md
```

`glossary.md` records confirmed terms or states that no project-specific terms are confirmed yet. `README.md` records current decisions and planned direction; `change-log.md` records when those decisions were added or changed.

Do not write current-state `modules/`, detailed walkthroughs, or API/schema/metrics module pages until there is code, config, data, or an explicit accepted architecture to support them. If the user asks for a planned architecture, mark every unimplemented module or walkthrough as `Planned`.

### Seed main guide

For seed projects, `repo-docs/README.md` should answer:

1. What is the project trying to become?
2. What is already present in the repo?
3. What has been explicitly decided?
4. What is the planned first workflow?
5. What are the next implementation steps?
6. What is unknown or waiting for verification?

Good seed README sections:

- Project brief
- Current repo state
- Decided constraints
- Planned first workflow
- Next implementation steps
- Unknowns and verification checks

### Seed next steps

Record planned implementation work in `README.md` (Next implementation steps) and explicit decisions in `README.md` plus `change-log.md`. Do not create a separate decision or change-plan page—the guide serves understanding, not maintenance routing.

### Seed change log

`change-log.md` should record project initialization and explicit user decisions:

| Timestamp | Request | Actions | Verification | Result |
| --- | --- | --- | --- | --- |

For seed projects, verification may be "docs scaffold created" or "decision recorded"; do not imply implementation verification when no implementation exists.

### Promotion rule

As code appears, promote seed content:

- `Planned` module -> module doc only after files or explicit architecture exist.
- `Planned` workflow -> current-state walkthrough or flow doc only after an entrypoint, script, or detailed accepted design exists. Until then, keep it in `README.md` as `Planned`.
- `Unknown` -> `Planned` when the user chooses a direction.
- `Planned` -> `Confirmed` only when repo evidence proves it.

## Specialized Repo Notes

For benchmark, evaluation, safety, data-heavy, or research repositories, make sure stable docs cover the relevant project facts without pretending every project is a paper. Simple tools and CRUD projects can skip research framing and explain the real design idea directly.

| Specialized concept | Typical location |
| --- | --- |
| Project question and motivation | `README.md` or main guide |
| Method / benchmark design | main guide, module docs |
| Experiment entrypoints | main guide, walkthroughs for multi-phase runs, entrypoint module |
| Metrics and evaluation protocol | evaluation or metrics module |
| Data/task/schema contracts | data, task, or schema module |
| Reproduction path | main guide, walkthroughs, scripts module when needed |
| Baselines / ablations / compared methods | method or experiment module |
| Output artifacts | artifacts or output module |
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: 472kUpdated Claude CodeCodexVerifiedInstall
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: 198kUpdated Claude CodeCodexVerifiedInstall
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: 198kUpdated Claude CodeCodexUnverifiedInstall
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: 198kUpdated Claude CodeCodexVerifiedInstall