MCPInstall

autonomous-common

Use when setting up, troubleshooting, or modifying the shared hooks and agent-callable utility scripts that enforce the autonomous dev/review workflow. Triggers on phrases like "push to main is blocked", "block-commit-outside-worktree hook failing", "configure hooks after npx skills add", "what does check-pr-review.sh do", "set up workflow hook symlinks", or when editing files under skills/autonomous-common/. Provides the hooks the autonomous-dev / autonomous-review skills depend on, plus utility scripts (gh-as-user.sh, mark-issue-checkbox.sh, reply-to-comments.sh, resolve-threads.sh).

Installs
98
Last 24h
—
Trending rank
—
Security
Not audited

Install

Works with every agent the skills CLI supports.

npx skills add https://github.com/zxkane/autonomous-dev-team --skill autonomous-common
  1. 01

    Install

    Run the command in your project. The CLI asks which agents to install for.

  2. 02

    Describe the task

    Ask your agent for something this skill covers. It loads autonomous-common automatically when relevant.

  3. 03

    Review the result

    The agent follows the SKILL.md steps, scripts and references bundled with the skill.

SKILL.md

The instructions your agent reads when the skill is triggered.

Autonomous Common Infrastructure

Shared workflow-enforcement hooks and agent-callable utility scripts used by autonomous-dev, autonomous-review, and autonomous-dispatcher. The other autonomous-* skills reference scripts and hooks here — when those reference paths break, this is usually the skill to look at.

For optional dev/review learning capture, scoped fact revision, local environment references, or post-merge documentation writeback, read references/session-knowledge.md. Assessments may report no update; only a useful documentation diff creates a commit/PR.

Setup for npx skills add Users

After npx skills add, run the installer for your coding agent once from the project root:

AgentInstallerWrites to
Claude Codebash .claude/skills/autonomous-common/scripts/install-claude-hooks.sh.claude/settings.json
Qoderbash .claude/skills/autonomous-common/scripts/install-qoder-hooks.sh.qoder/settings.json
Antigravitybash .claude/skills/autonomous-common/scripts/install-antigravity-hooks.sh.antigravity/hooks.json
Cursorbash .claude/skills/autonomous-common/scripts/install-cursor-hooks.sh.cursor/hooks.json
Kiro CLIbash .claude/skills/autonomous-common/scripts/install-kiro-hooks.sh [--agent <name>].kiro/agents/<name>.json (default: default)
Gemini CLIbash .claude/skills/autonomous-common/scripts/install-gemini-hooks.sh.gemini/settings.json
Codex CLIbash .claude/skills/autonomous-common/scripts/install-codex-hooks.sh (Python 3.11+).codex/hooks.json + .codex/config.toml
Windsurfbash .claude/skills/autonomous-common/scripts/install-windsurf-hooks.sh.windsurf/hooks.json
Kimi CLIbash .claude/skills/autonomous-common/scripts/install-kimi-hooks.sh [--project]~/.kimi/config.toml (default; --project writes .kimi/config.toml)

Each installer wires up the workflow hooks at the project scope, so they fire on every shell command in the repo — not only when an autonomous-* skill is explicitly loaded. Without this, the hook commands declared in skill frontmatter only run while a skill is active in the conversation, which is the regression that closed #68.

For the per-agent schema mapping reference, see docs/cross-agent-hooks.md.

Project-side scripts/ and hooks/ symlinks

The IDE installer above only writes the IDE config file (e.g. .claude/settings.json). The project-side <project>/scripts/ symlinks (so dispatcher-tick.sh finds autonomous-dev.sh, lib-agent.sh, etc.) and the <project>/hooks directory symlink are managed by a separate, IDE-agnostic bootstrap script — the canonical pattern for projects whose scripts/ already contains project-local files:

# From your project root, after `npx skills add`:
bash .agents/skills/autonomous-common/scripts/install-project-hooks.sh

What it does:

  • Symlinks every *.sh from the installed autonomous-dispatcher/scripts/ into <project>/scripts/, without overwriting real (non-symlink) project-local files like autonomous.conf or per-project deploy helpers.
  • Prunes dangling symlinks if upstream removes a file.
  • Symlinks <project>/hooks → autonomous-common/hooks (refuses to shadow an existing real directory).
  • Installs the per-worktree git pre-push hook (#65). Skip with --no-git-hook.

Idempotent — re-run after every npx skills update so newly-added upstream files (e.g. when this skill set adds a new lib-*.sh) are picked up automatically. Closes the silent-drift mode behind #153, where projects bootstrapped via per-file ln -s cargo-culted lists missed new files and autonomous-review.sh died on source of the missing file before any review work ran.

Legacy directory-level fallback (deprecated)

The earlier docs suggested replacing the project's scripts/ directory with a single symlink:

ln -sf .claude/skills/autonomous-dispatcher/scripts scripts   # DEPRECATED

This loses any project-local files in scripts/. Use install-project-hooks.sh instead — it does the right thing on a directory that already has project-local content, and re-running picks up upstream changes.

Required Claude Code plugins

Claude Code only. The installer prompts for these; if installing manually, add to .claude/settings.json under enabledPlugins:

{
  "enabledPlugins": {
    "code-simplifier@claude-plugins-official": true,
    "pr-review-toolkit@claude-plugins-official": true
  }
}

IDEs without hook support (Cursor, Windsurf) skip both the installer and the symlinks — the skills work without hooks, but workflow steps must be followed manually.

What's here

  • hooks/ — workflow-enforcement hooks (block-push-to-main, block-commit-outside-worktree, check-pr-review, check-shellcheck, verify-completion, …). See hooks/README.md for the canonical list and per-hook semantics.
  • scripts/ — agent-callable utilities used by the dev/review skills:
    • session-knowledge.sh / session-knowledge.py — private shared task receipts, optional verified candidates, local knowledge, and isolated documentation commits (Python standard library)
    • install-project-hooks.sh — IDE-agnostic project-side bootstrap: symlinks dispatcher *.sh into <project>/scripts/ (without overwriting project-local files), symlinks <project>/hooks, prunes dangling links, installs the git pre-push hook. Re-run after every npx skills update (closes #153)
    • lib-installer.sh — shared merge/write helpers used by every per-agent installer
    • lib-installer-translate.sh — schema translation helpers for near-clone agents (event-name map, tool-name map, timeout-unit conversion)
    • install-claude-hooks.sh — Claude Code installer (writes .claude/settings.json)
    • install-qoder-hooks.sh — Qoder installer (writes .qoder/settings.json — same schema as Claude Code)
    • install-antigravity-hooks.sh — Antigravity installer (writes .antigravity/hooks.json — hooks-only file; contract is community-observed, undocumented by Google)
    • install-cursor-hooks.sh — Cursor installer (writes .cursor/hooks.json — version: 1 envelope, camelCase events, Shell matcher)
    • install-kiro-hooks.sh — Kiro CLI / Amazon Q installer (writes .kiro/agents/<name>.json — agent definition with camelCase events, execute_bash/fs_write matchers, timeout_ms in milliseconds)
    • install-gemini-hooks.sh — Gemini CLI installer (writes .gemini/settings.json — BeforeTool/AfterTool events, run_shell_command/write_file/replace matchers)
    • install-codex-hooks.sh — Codex CLI installer (requires Python 3.11+ tomllib, renders worktree-root-resolvable .codex/hooks.json, sets canonical [features] hooks = true, and safely migrates the deprecated codex_hooks alias)
    • install-windsurf-hooks.sh — Windsurf installer (writes .windsurf/hooks.json — snake_case events that fold matcher info; no per-tool matcher field)
    • install-kimi-hooks.sh — Kimi CLI installer (writes ~/.kimi/config.toml user-level by default, or .kimi/config.toml with --project; emits TOML [[hooks]] blocks)
    • claude-settings.template.json — canonical hook list applied by all per-agent installers
    • gh-as-user.sh — runs gh as a real user (needed when retriggering bot reviews like /q review)
    • mark-issue-checkbox.sh — toggles GitHub issue body checkboxes from the agent
    • reply-to-comments.sh — replies to PR review comments
    • resolve-threads.sh — batch-resolves review threads on a PR

The hooks and scripts are documented in detail in their respective README/source files. This SKILL.md only catalogs what's available so you can find the right file to edit.

FAQ

What does the autonomous-common skill do?

Use when setting up, troubleshooting, or modifying the shared hooks and agent-callable utility scripts that enforce the autonomous dev/review workflow. Triggers on phrases like "push to main is blocked", "block-commit-outside-worktree hook failing", "configure hooks after npx skills add", "what does check-pr-review.sh do", "set up workflow hook symlinks", or when editing files under skills/autonomous-common/. Provides the hooks the autonomous-dev / autonomous-review skills depend on, plus utility scripts (gh-as-user.sh, mark-issue-checkbox.sh, reply-to-comments.sh, resolve-threads.sh).

How do I install autonomous-common?

Run npx skills add https://github.com/zxkane/autonomous-dev-team --skill autonomous-common in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/autonomous-common/ (personal) or .claude/skills/autonomous-common/ (project).

Which AI agents can use autonomous-common?

Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.

Is autonomous-common safe to install?

No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.

Who maintains autonomous-common?

It is published by zxkane in zxkane/autonomous-dev-team. It has 98 recorded installs.

More from zxkane/autonomous-dev-team

View all →

autonomous-dispatcher

zxkane/autonomous-dev-team

Open the skill to read its SKILL.md instructions.

124 installs

create-issue

zxkane/autonomous-dev-team

Use when the user asks to create a GitHub issue, file a bug, request a feature, open a tracking issue, or break a feature into multiple sub-issues. Guides interactive issue drafting with structured templates, workspace-change attachment, dependency linking, and the optional autonomous label for the automated dev pipeline.

autonomous-review

zxkane/autonomous-dev-team

Use to perform an end-to-end PR review and reach an approve/request-changes verdict — including verifying acceptance criteria, running E2E tests via browser automation, resolving merge conflicts, and (when verdict passes) merging the PR. Triggers on phrases like "review this PR", "decide whether to approve and merge", "run E2E verification", "resolve merge conflicts on PR N", or when the dispatcher hands off a PR labeled pending-review / reviewing for autonomous review. Distinct from in-flight dev-side self-review (that lives in autonomous-dev's pr-review step).

autonomous-dev

zxkane/autonomous-dev-team

Use to develop a feature or bug fix end-to-end through a TDD git-worktree workflow — interactively (developer-led) or unattended (autonomous-mode, driven by the dispatcher). Triggers on phrases like "implement issue N", "fix this bug", "add a feature", "create a worktree", "write test cases", "push and open a PR", "check CI", "address review comments", "resolve review threads", "/q review", "/codex review", "implement this autonomously", or any partial step in the design → worktree → tests → implement → verify → review → PR → CI → E2E lifecycle. Interactive mode asks for decisions; autonomous mode makes decisions per autonomous-mode.md and posts progress comments to the GitHub issue.

Popular code quality & workflow skills

View category →

grill-me

mattpocock/skills

A relentless interview to sharpen a plan or design.

Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.

handoff

mattpocock/skills

Compact the current conversation into a handoff document for another agent to pick up.

domain-modeling

mattpocock/skills

Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing a GLOSSARY.md, or recording or editing an ADR.

codebase-design

mattpocock/skills

Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.

diagnosing-bugs

mattpocock/skills

Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.