{
  "markdown": "# claude-brewcode\n\n[![macOS](https://img.shields.io/badge/macOS-support-000000?logo=apple&logoColor=white)](https://www.apple.com/macos/)\n[![Claude](https://img.shields.io/badge/Claude-Anthropic-orange?logo=anthropic&logoColor=white)](https://claude.ai)\n[![Claude Code](https://img.shields.io/badge/Claude_Code-Plugin-blueviolet?logo=anthropic&logoColor=white)](https://code.claude.com)\n\n[![Release](https://github.com/kochetkov-ma/claude-brewcode/actions/workflows/release.yml/badge.svg)](https://github.com/kochetkov-ma/claude-brewcode/actions/workflows/release.yml)\n[![Latest Release](https://img.shields.io/github/v/release/kochetkov-ma/claude-brewcode?label=latest&color=blue)](https://github.com/kochetkov-ma/claude-brewcode/releases/latest)\n[![Documentation](https://img.shields.io/badge/Docs-doc--claude.brewcode.app-4A90D9?logo=bookstack&logoColor=white)](https://doc-claude.brewcode.app/getting-started/)\n\n**Claude Code plugin suite** -- four plugins for development, documentation, text utility, and visual workflows.\n\nA regular Claude Code session hands a big task to one agent and loses sight of it. Brewcode splits work into bounded units, gives every spawn a six-field brief, and re-states the delegation rule on every prompt. Four plugins. 28 skills. 8 agents. 6 lifecycle hooks.\n\n[**Full Documentation**](https://doc-claude.brewcode.app/getting-started/)\n\n[**Approaches**](APPROACHES.md) -- every hook, agent and injection the suite installs, and who installs what.\n\n> **Security:** Code is scanned on every commit. No personal data is collected or transmitted.\n>\n> [![Gitleaks](https://github.com/kochetkov-ma/claude-brewcode/actions/workflows/gitleaks.yml/badge.svg)](https://github.com/kochetkov-ma/claude-brewcode/actions/workflows/gitleaks.yml?query=branch%3Amain)\n> [![CodeQL](https://github.com/kochetkov-ma/claude-brewcode/actions/workflows/codeql.yml/badge.svg)](https://github.com/kochetkov-ma/claude-brewcode/actions/workflows/codeql.yml?query=branch%3Amain)\n> [![Semgrep](https://github.com/kochetkov-ma/claude-brewcode/actions/workflows/semgrep.yml/badge.svg)](https://github.com/kochetkov-ma/claude-brewcode/actions/workflows/semgrep.yml?query=branch%3Amain)\n\n## 🚀 Install in 30 seconds\n\nPaste this prompt into any Claude Code session -- Claude will run the full install for you:\n\n```\nExecute these commands in this Claude Code session, one by one, show full output for each, do not skip any, do not summarize:\n\n1. claude plugin marketplace add https://github.com/kochetkov-ma/claude-brewcode\n2. claude plugin install brewcode@claude-brewcode\n3. claude plugin install brewdoc@claude-brewcode\n4. claude plugin install brewtools@claude-brewcode\n5. claude plugin install brewui@claude-brewcode\n\nAfter all commands succeed, run `/reload-plugins`. If `/reload-plugins` is unavailable, tell me to type `exit` and run `claude` again. Run the commands now.\n```\n\nAlready installed? Update with this prompt:\n\n```\nExecute these commands in this Claude Code session, one by one, show full output for each, do not skip any, do not summarize:\n\n1. claude plugin marketplace update claude-brewcode\n2. claude plugin update brewcode@claude-brewcode\n3. claude plugin update brewdoc@claude-brewcode\n4. claude plugin update brewtools@claude-brewcode\n5. claude plugin update brewui@claude-brewcode\n\nAfter all commands succeed, run `/reload-plugins`. If `/reload-plugins` is unavailable, tell me to type `exit` and run `claude` again. Run the commands now.\n```\n\n> After install/update, run `/reload-plugins` (or `exit` + `claude` to restart). For manual install steps see [Installation](#installation) below.\n\n## Plugin Suite\n\n| Plugin | Purpose | Skills | Install |\n|--------|---------|--------|---------|\n| [brewcode](brewcode/README.md) | Infinite task execution, quorum reviews, skill/agent creation, semantic search | 9 | `claude plugin install brewcode@claude-brewcode` |\n| [brewdoc](brewdoc/README.md) | Documentation tools: docsync, memory-sync generation, PDF conversion, publishing | 5 | `claude plugin install brewdoc@claude-brewcode` |\n| [brewtools](brewtools/README.md) | Universal text utilities: token optimization, humanization, secrets scanning, plugin updates | 14 | `claude plugin install brewtools@claude-brewcode` |\n| [brewui](brewui/README.md) | UI/visual/creative tools (placeholder, currently empty) | 0 | `claude plugin install brewui@claude-brewcode` |\n\n## Installation\n\n### Marketplace (recommended)\n\nPermanent installation through the Claude Code plugin system. Plugins load automatically with every session.\n\n1. **Add marketplace** -- registers the plugin source:\n\n```bash\nclaude plugin marketplace add https://github.com/kochetkov-ma/claude-brewcode\n```\n\n2. **Install plugins** -- copies plugins to local cache:\n\n```bash\nclaude plugin install brewcode@claude-brewcode\nclaude plugin install brewdoc@claude-brewcode\nclaude plugin install brewtools@claude-brewcode\nclaude plugin install brewui@claude-brewcode\n```\n\n3. **Reload plugins** -- run `/reload-plugins` in Claude Code, or `exit` + `claude` to restart.\n\n### Already installed? Update\n\nUse `/brewtools:plugin-update` inside Claude Code for the easiest path -- it runs all update commands and reloads automatically.\n\nOr run manually:\n\n```bash\nclaude plugin marketplace update claude-brewcode\nclaude plugin update brewcode@claude-brewcode\nclaude plugin update brewdoc@claude-brewcode\nclaude plugin update brewtools@claude-brewcode\nclaude plugin update brewui@claude-brewcode\n```\n\nAfter updating, run `/reload-plugins` (preferred) or `exit` + `claude` to restart.\n\n### Local directory (for development)\n\n```bash\nclaude --plugin-dir ./brewcode --plugin-dir ./brewdoc --plugin-dir ./brewtools --plugin-dir ./brewui\n```\n\n### Requirements\n\n| Component | Version | Purpose |\n|-----------|---------|---------|\n| Claude Code CLI | latest | Plugin runtime |\n| Node.js | 20+ | Hook runtime |\n| macOS | 13+ | Supported OS |\n\n> **macOS only.** Linux and Windows support is planned for future releases.\n\n## Quick Start\n\n### brewcode -- infinite task execution\n\n```bash\n/brewtools:task-board-setup  # 1. Deploy the task board + a project-tailored /task-spec skill\n/task-spec \"Implement JWT authorization\"  # 2. Research codebase + write the task spec and design\n/brewcode:superreview-setup  # 3. Generate a project-tailored deep-review skill\n/brewcode:setup-status       # anytime: what is installed, stale or missing in this project\n```\n\nSkills orchestrate, agents execute. Each spawn is a bounded unit with a six-field brief; the `forced-eval` hook re-states the manager role and the split rule on every prompt, and a `role-recall`/`compact-recall` pair re-anchors the same role plus the plan and task graph after a compaction, so work stays observable across compaction cycles.\n\n### brewdoc -- documentation tools\n\n```bash\n/brewdoc:docsync-setup                # Install doc-staleness tracking, then sync stale docs\n/brewdoc:my-claude                    # Generate Claude Code installation docs\n/brewdoc:memory-sync-setup            # Generate a project-tailored /memory-sync skill into the repo\n/brewdoc:md-to-pdf ./docs/report.md   # Convert markdown to PDF\n/brewdoc:publish \"Hello world\"        # Publish to brewpage.app -- returns public URL\n```\n\n### brewtools -- text utilities\n\n```bash\n/brewtools:text-optimize CLAUDE.md         # Token-efficient optimization (52 rules)\n/brewtools:text-human 3be67487             # Remove AI artifacts from a commit\n/brewtools:secrets-scan                    # Scan for leaked credentials\n/brewtools:plugin-update                   # Install or update the plugin suite\n/brewtools:context-slim measure            # Weigh the permanent context surface, read-only\n```\n\n### brewui -- visual tools\n\nPlaceholder plugin, currently empty. No commands yet -- coming soon.\n\n## How It Works\n\n```\n  /task-spec \"...\" --> parallel research agents + user Q&A --> per-task spec + design docs\n        │\n        v\n  project agents from .claude/agents/ --> bounded units, fanned out in ONE message\n        │\n        v\n  /brewcode:superreview-setup --> project-tailored deep-review skill\n\n  every prompt: forced-eval (UserPromptSubmit) injects 3 lines --\n    [ROLE]   scan agents, project .claude/agents/ first; domain expert exists -> delegate\n    [SPLIT]  one agent for an hour = drift you cannot observe -> split and fan out\n    [BRANCH] no branch chosen -> main; no explicit branch/PR -> take over the whole workspace\n\n  after a compaction: role-recall + compact-recall (SessionStart, matcher compact) re-inject\n    the same 3 lines plus plan/intent and task graph -- forced-eval never fires there, since\n    an auto-compaction has no prompt to hook into\n```\n\n### Delegation contract\n\nEvery skill that spawns subagents carries a `## Delegation` section; every shipped agent carries a `## Scope guard`. Both encode the same rule: one subagent = ONE bounded unit -- one deliverable, ~5 files, ~10 steps. Anything bigger is split into N tasks spawned in a single message.\n\nEvery spawn prompt carries six fields:\n\n| Field | Content |\n|-------|---------|\n| GOAL | the overall task and why it exists -- the point beyond the file edit |\n| ROLE | what this agent owns; what it must NOT touch |\n| SCOPE | exact paths and commands in bounds, plus explicit out-of-bounds |\n| CONTEXT | what is already done, by whom, what runs in parallel |\n| CONSUMER | who or what uses the result next, and the shape it must fit |\n| DONE | acceptance criteria plus the exact report shape expected back |\n\n## Skills Reference\n\n> **The `-setup` suffix** marks a skill you run once to install a mechanism -- afterwards you use what it produced (a generated skill, a hook, an MCP server), not the skill itself. Recurring tools you invoke every day keep bare names. Every `-setup` skill shares one mode vocabulary: `status | install | upgrade | enable | disable | uninstall | purge`, and no argument means `status` when installed, `install` when not. The one exception is `/brewcode:semble-setup`, which always defaults to `status` so a bare invocation never triggers a machine-level package install. All eleven setups implement all seven verbs, via one of two mechanisms: a live config flag re-checked each invocation (semble, agent-deadline, agent-return, agent-router, manager, docsync), or entry-file parking, where the filename discovery keys on is renamed `<name>.disabled` with the body byte-identical (teams, superreview, task-board, think-short, memory-sync).\n\n> **All 28 skills are user-invoked only.** Every one carries `user-invocable: true` **and** `disable-model-invocation: true` in its frontmatter: the model never sees their descriptions and never auto-activates one. You type `/plugin:skill`, or nothing runs. This is a deliberate trade about context cost -- 28 model-visible descriptions would be a permanent tax on every request -- and these skills do not want auto-activation anyway: eleven of them write real files into your repo after asking you real questions, and the rest are tools you point at a scope you choose.\n\n### Brewcode (9 skills)\n\n| Skill | Purpose |\n|-------|---------|\n| `/brewcode:setup-status` | Read-only cross-plugin dashboard: which setup skills are installed, stale, disabled, partial or missing here, plus the exact command to run for each. `disabled` outranks `partial`/`stale`, so a mechanism you turned off on purpose is never reported as broken. Runs nothing itself -- setups are interactive generators that spawn many subagents, and stacking several in one session degrades all of them |\n| `/brewcode:superreview-setup` | Generate a project-tailored deep-review skill: `QUICK` (default, `intent-guard` + mechanical gates) or `EXTENDED` (adds domain-expert fan-out, scope discipline, adversarial validation) depth, read from your prompt |\n| `/brewcode:teams-setup` | Create and manage dynamic teams of domain-specific agents -- every team also gets a fixed review-only `intent-guard` member (not counted in team size). Modes `status`/`install`/`upgrade`/`enable`/`disable`/`uninstall`/`purge`, each taking an optional team `[name]`; `enable`/`disable` park/unpark each roster member's agent file |\n| `/brewcode:convention` | Extract etalon classes, patterns, architecture into convention docs |\n| `/brewcode:rules` | Prompt-driven rules management: status, create, improve, review |\n| `/brewcode:skills` | Prompt-driven skill management: status, create, improve, sync, review |\n| `/brewcode:agents` | Prompt-driven agent management: status, create, improve, sync, review |\n| `/brewcode:e2e` | E2E testing orchestration with BDD scenarios and quorum review |\n| `/brewcode:semble-setup` | Semantic code search setup: installs the pinned semble_code MCP, shared content-variant cache, semble-first rule + hooks, agent migration |\n\n### Brewdoc (5 skills)\n\n| Skill | Purpose |\n|-------|---------|\n| `/brewdoc:docsync-setup` | Install project-local doc-staleness tracking (hooks + config), then report or force a sync |\n| `/brewdoc:my-claude` | Generate Claude Code installation docs |\n| `/brewdoc:memory-sync-setup` | Generate a project-tailored `/memory-sync` skill: syncs everything auto-loaded into context (root & nested CLAUDE.md, CLAUDE.local.md, rules, conventions, AGENTS.md family, agents, skills, memory dir) against the code -- docs excluded; scopes `session` (default), `branch`, `commit <sha>`, `recent[:N]`, `all`, plus a `hard` depth (rules `paths:` precision audit + obvious-knowledge purge); non-growth, agents re-audited every run |\n| `/brewdoc:md-to-pdf` | Convert markdown to professional PDF |\n| `/brewdoc:publish` | Publish to brewpage.app -- returns public URL |\n\n### Brewtools (14 skills)\n\n| Skill | Purpose |\n|-------|---------|\n| `/brewtools:text-optimize` | LLM token efficiency optimization (52 rules, smart dedup + aggressive lossy) |\n| `/brewtools:text-human` | Remove AI artifacts, humanize code |\n| `/brewtools:think-short-setup` | Install/remove terse-mode hooks (SessionStart + every-10th UserPromptSubmit + SubagentStart) that inject brevity directives; project or global |\n| `/brewtools:agent-deadline-setup` | Install/remove a soft wall-clock budget for subagents -- 80% warns \"wrap up\", 100% blocks all but finalization tools; project or global, opt-in |\n| `/brewtools:agent-return-setup` | Install/remove a size budget on every subagent's final return -- a SubagentStart hook states the budget, a SubagentStop hook sizes the return (`chars/4`) and blocks at most once, ordering a compress above `passTokens` (default 1000) or a write-to-file above `fileTokens` (default 2500); project or global, opt-in |\n| `/brewtools:agent-router-setup` | EXPERIMENTAL -- install/remove a PreToolUse hook that denies a generic subagent spawn in favor of the real project/plugin expert, or nudges when the fit is only uncertain; project scope only, opt-in |\n| `/brewtools:secrets-scan` | Scan git-tracked files for leaked secrets |\n| `/brewtools:ssh` | SSH server management -- connect, configure, deploy |\n| `/brewtools:deploy` | GitHub Actions deployment -- workflows, releases, GHCR, CI/CD |\n| `/brewtools:plugin-update` | Install and update the full plugin suite |\n| `/brewtools:provider-switch` | Configure alternative API providers (DeepSeek, Z.ai/GLM, Qwen, MiniMax, OpenRouter) |\n| `/brewtools:manager-setup` | Manager mode -- hook-driven codewords `++m` (delegate-everything, plan-aware), `++a` (architecture-first), `++rr` (anti-regression review), `++r` (two-phase double-check); the opt-in HARD wall (`status`/`install`/`upgrade`/`enable`/`disable`/`uninstall`/`purge`, plus `level strict\\|balanced` and `edit`) installs a project PreToolUse guard that blocks main-session writes while subagents stay free -- an allowlist of permitted binaries and per-binary flags, failing closed when it cannot parse the command |\n| `/brewtools:task-board-setup` | Deploy a file-based Kanban into ANY repo via multi-agent analysis -- task-tracker agent, task-board skill, tasks rule, .claude/features, plus an optional spec + design layer (`task-spec`) and an `upgrade` mode for existing boards |\n| `/brewtools:context-slim` | Shrink the permanent context surface (project+global CLAUDE.md, rules, agent descriptions, hooks, memory) via cross-layer dedup, default-knowledge removal and per-file compression -- lossless by default, whole-run rollback on any verify miss |\n\n### Brewui (0 skills)\n\nNo skills yet -- placeholder for future UI/visual/creative tools.\n\n### Portable skills (standalone)\n\nSelf-contained `SKILL.md` folders that ship outside the four plugins -- drop them into any compatible agent runtime. Both publish content to [brewpage.app](https://brewpage.app) (text, markdown, files, multi-file sites) and return a public URL.\n\n| Skill | Runtime | Path |\n|-------|---------|------|\n| `brewpage-publish` | Claude Code | [`skills/brewpage-publish`](skills/brewpage-publish/) |\n| `brewpage-publish` | OpenClaw / AgentSkills | [`openclaw/brewpage-publish`](openclaw/brewpage-publish/) |\n\n## Agents (8 total)\n\n| Agent | Plugin | Model | Purpose |\n|-------|--------|-------|---------|\n| skill-creator | brewcode | inherit | Create and improve Claude Code skills |\n| agent-creator | brewcode | inherit | Create and improve Claude Code agents |\n| hook-creator | brewcode | inherit | Create and debug Claude Code hooks |\n| bash-expert | brewcode | inherit | Create professional shell scripts |\n| bc-rules-organizer | brewcode | haiku | Internal: spawned by /brewcode:rules |\n| text-optimizer | brewtools | sonnet | Optimize text and docs for LLM efficiency |\n| ssh-admin | brewtools | inherit | Linux server administration via SSH |\n| deploy-admin | brewtools | inherit | GitHub Actions deployment and CI/CD |\n\n> **Scope guard:** every shipped agent carries a `## Scope guard` -- a task larger than one bounded unit (one deliverable, ~5 files, ~10 steps) makes the agent stop before starting and return a split proposal instead of grinding for an hour. `ssh-admin` and `deploy-admin` additionally split per host, repo, and environment.\n\n## Documentation\n\n| Resource | Link |\n|----------|------|\n| Full documentation | [doc-claude.brewcode.app](https://doc-claude.brewcode.app/getting-started/) |\n| Quick Start guide | [Quickstart](https://doc-claude.brewcode.app/quickstart/) |\n| Brewcode overview | [Brewcode](https://doc-claude.brewcode.app/brewcode/overview/) |\n| Brewdoc overview | [Brewdoc](https://doc-claude.brewcode.app/brewdoc/overview/) |\n| Brewtools overview | [Brewtools](https://doc-claude.brewcode.app/brewtools/overview/) |\n| Brewui overview | [Brewui](https://doc-claude.brewcode.app/brewui/overview/) |\n| Release Notes | [RELEASE-NOTES.md](RELEASE-NOTES.md) |\n\n## Development\n\n```bash\nCLAUDE_DEBUG=1 claude --plugin-dir ./brewcode   # Debug mode\nbash .claude/scripts/update-plugin.sh           # Update all plugins\nbash .claude/scripts/bump-version.sh X.Y.Z      # Bump version everywhere\n```\n\n## License\n\nMIT -- see [LICENSE](LICENSE)\n",
  "bytes": 18894,
  "sha": "9ef702cec0bdca6fec77779d18b9571acafcfb6cb4d9de03daadfae70a19e72d",
  "repo_slug": "kochetkov-ma/claude-brewcode",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/skl_kochetkov_ma_claude_brewcode_text_optimi_9d2ab0ba/readme"
}