/docs
Update repository documentation to match the current state of the codebase. Local replacement for the remote /docs command (which needs the Claude GitHub app). Use when the user says /docs, "update the docs", "sync the README", "document this feature", or asks you to refresh
$ npx -y skills add ReinaMacCredy/maestro --skill docs --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
- Slash command
/docs
Context preview
The summary Claude sees to decide when to auto-load this skill.
Update repository documentation to match the current state of the codebase. Local replacement for the remote /docs command (which needs the Claude GitHub app). Use when the user says /docs, "update the docs", "sync the README", "document this feature", or asks you to refresh
SKILL.md
docs.SKILL.mdname: docs
description: >-
Update repository documentation to match the current state of the codebase.
Local replacement for the remote /docs command (which needs the Claude GitHub
app). Use when the user says /docs, "update the docs", "sync the README",
"document this feature", or asks you to refresh docs to reflect current code.
Works inline in the current session and edits files in the working tree.
docs
Local, in-session documentation updater. Replaces the remote `/docs` command so you do not need the Claude GitHub app or a cloud session. No PR is opened. Changes are left staged in the working tree for the user to review and commit.
Scope the task
Read the user's request literally before touching anything.
- Full repo sync ("update the docs", "sync README"): update `README.md` and relevant `docs/*.md`. Prefer delegating the mechanical work to the `doc-sync` skill if it is available in this session.
- Specific file ("update AGENTS.md", "refresh CLAUDE.md"): edit only that file.
- Feature doc ("document the task feature"): extend the existing doc for that feature. Create a new file in `docs/` only if the user explicitly asked for a new file.
- No argument / vague ask: ask one direct question about which file or area to update, then stop.
Ground every claim in current state
Never document from memory. Every factual claim you write (file path, function name, flag, command, version, behavior) must trace to something you just read or ran.
Required reads before writing:
- `AGENTS.md` and `CLAUDE.md` at the repo root: source of truth for conventions, style, and project rules.
- For feature docs: that feature's `index.ts`, `services.ts`, and `commands/` to see the real public surface.
- For CLI docs: run the binary's `--help` to capture real flags and argument order. Do not invent flags.
- For version or release docs: check `package.json` and `git log --oneline -20` for recent commits.
If you cannot verify a claim right now, mark it `(unverified)` inline or delete it. Do not ship an unverified specific.
Edit surgically
- Prefer editing existing files over creating new ones.
- Match the existing heading style, tone, and voice.
- Do not reformat adjacent content that is not part of your change.
- Do not "improve" code or comments you happen to read while gathering context. That is a separate task.
- Plain markers only: `[!]`, `[ok]`, `-->`, `[x]`. No emoji. No em-dashes. Project rule from CLAUDE.md.
Verify after writing
After each file edit:
- Re-read the changed section against the file or command output you read.
- If you claimed a command exists, run it with `--help` and confirm the flags.
- If you referenced a path, confirm the file exists with `Glob` or `Read`.
Do not commit
Leave changes staged in the working tree. The user reviews and commits.
If the user explicitly asks for a commit, use Conventional Commits: `docs(scope): short summary`. Do not mix doc changes with unrelated behavior changes in one commit.
Guardrails
- Do not touch `.maestro/memory/**`, `.maestro/tasks/**`, or `.maestro/missions/**`.
- Do not touch files under `~/.claude/projects/**/memory/**`.
- Read-only tools (`Read`, `Grep`, `Glob`, `git log`, `--help`) are always fine. Any mutation outside doc files requires explicit user authorization.
- Never push, force-push, amend, or rewrite history as part of a docs task.
Report
End with a brief report, at most 6 bullets:
- Files changed, absolute paths.
- One line per file: what changed and why.
- Anything flagged `(unverified)` and why you could not verify it.
- Suggested commit message if the user wants to commit.
Read more
name: docs description: >- Update repository documentation to match the current state of the codebase. Local replacement for the remote /docs command (which needs the Claude GitHub app). Use when the user says /docs, "update the docs", "sync the README", "document this feature", or asks you to refresh docs to reflect current code. Works inline in the current session and edits files in the working tree.
docs
Local, in-session documentation updater. Replaces the remote `/docs` command so you do not need the Claude GitHub app or a cloud session. No PR is opened. Changes are left staged in the working tree for the user to review and commit.
Scope the task
Read the user's request literally before touching anything.
- Full repo sync ("update the docs", "sync README"): update `README.md` and relevant `docs/*.md`. Prefer delegating the mechanical work to the `doc-sync` skill if it is available in this session.
- Specific file ("update AGENTS.md", "refresh CLAUDE.md"): edit only that file.
- Feature doc ("document the task feature"): extend the existing doc for that feature. Create a new file in `docs/` only if the user explicitly asked for a new file.
- No argument / vague ask: ask one direct question about which file or area to update, then stop.
Ground every claim in current state
Never document from memory. Every factual claim you write (file path, function name, flag, command, version, behavior) must trace to something you just read or ran.
Required reads before writing:
- `AGENTS.md` and `CLAUDE.md` at the repo root: source of truth for conventions, style, and project rules.
- For feature docs: that feature's `index.ts`, `services.ts`, and `commands/` to see the real public surface.
- For CLI docs: run the binary's `--help` to capture real flags and argument order. Do not invent flags.
- For version or release docs: check `package.json` and `git log --oneline -20` for recent commits.
If you cannot verify a claim right now, mark it `(unverified)` inline or delete it. Do not ship an unverified specific.
Edit surgically
- Prefer editing existing files over creating new ones.
- Match the existing heading style, tone, and voice.
- Do not reformat adjacent content that is not part of your change.
- Do not "improve" code or comments you happen to read while gathering context. That is a separate task.
- Plain markers only: `[!]`, `[ok]`, `-->`, `[x]`. No emoji. No em-dashes. Project rule from CLAUDE.md.
Verify after writing
After each file edit:
- Re-read the changed section against the file or command output you read.
- If you claimed a command exists, run it with `--help` and confirm the flags.
- If you referenced a path, confirm the file exists with `Glob` or `Read`.
Do not commit
Leave changes staged in the working tree. The user reviews and commits.
If the user explicitly asks for a commit, use Conventional Commits: `docs(scope): short summary`. Do not mix doc changes with unrelated behavior changes in one commit.
Guardrails
- Do not touch `.maestro/memory/**`, `.maestro/tasks/**`, or `.maestro/missions/**`.
- Do not touch files under `~/.claude/projects/**/memory/**`.
- Read-only tools (`Read`, `Grep`, `Glob`, `git log`, `--help`) are always fine. Any mutation outside doc files requires explicit user authorization.
- Never push, force-push, amend, or rewrite history as part of a docs task.
Report
End with a brief report, at most 6 bullets:
- Files changed, absolute paths.
- One line per file: what changed and why.
- Anything flagged `(unverified)` and why you could not verify it.
- Suggested commit message if the user wants to commit.
Local-first harness for agent-built codebases. Humans steer, agents execute, maestro is the substrate. maestro is a single Rust binary that gives a coding agent a durable place to work.
Repo: ReinaMacCredy/maestro
Other skills on maestro.
- /cli-for-agent
Designs or reviews CLIs so coding agents can run them reliably: non-interactive flags, layered --help with examples, stdin/pipelines, fast actionable errors, idempotency, dry-run, and predictable structure. Use when building a CLI, adding commands, writing --help, or when the
Open skill - /gitnexus-cli
Use when the user needs to run GitNexus CLI commands like analyze/index a repo, check status, clean the index, generate a wiki, or list indexed repos. Examples: \"Index this repo\", \"Reanalyze the codebase\", \"Generate a wiki\"
Open skill - /gitnexus-debugging
Use when the user is debugging a bug, tracing an error, or asking why something fails. Examples: \"Why is X failing?\", \"Where does this error come from?\", \"Trace this bug\"
Open skill - /gitnexus-exploring
Use when the user asks how code works, wants to understand architecture, trace execution flows, or explore unfamiliar parts of the codebase. Examples: \"How does X work?\", \"What calls this function?\", \"Show me the auth flow\"
Open skill - /gitnexus-guide
Use when the user asks about GitNexus itself — available tools, how to query the knowledge graph, MCP resources, graph schema, or workflow reference. Examples: \"What GitNexus tools are available?\", \"How do I use GitNexus?\"
Open skill - /gitnexus-impact-analysis
Use when the user wants to know what will break if they change something, or needs safety analysis before editing code. Examples: \"Is it safe to change X?\", \"What depends on this?\", \"What will break?\"
Open skill

