Brewtools Overview

Brewtools is a lightweight plugin providing universal utilities: token-efficient optimization, AI-artifact removal, secrets scanning, SSH server management, and GitHub Actions deployment. Each skill is self-contained and requires no prior setup.

13 skills · 3 agents · 2 registered plugin hook commands. Setup skills install their optional project or global mechanisms separately.

Installation

Execute these commands in this Claude Code session, one by one, show full output for each, do not skip any, do not summarize:

1. claude plugin marketplace add https://github.com/kochetkov-ma/claude-brewcode
2. claude plugin install brewtools@claude-brewcode

After all commands succeed, run `/reload-plugins`. If `/reload-plugins` is unavailable, tell me to type `exit` and run `claude` again. Run the commands now.

Install the entire suite:

Execute these commands in this Claude Code session, one by one, show full output for each, do not skip any, do not summarize:

1. claude plugin marketplace add https://github.com/kochetkov-ma/claude-brewcode
2. claude plugin install brewcode@claude-brewcode
3. claude plugin install brewdoc@claude-brewcode
4. claude plugin install brewtools@claude-brewcode
5. claude plugin install brewui@claude-brewcode

After all commands succeed, run `/reload-plugins`. If `/reload-plugins` is unavailable, tell me to type `exit` and run `claude` again. Run the commands now.

Skills (recipes)

Commands

Task methodology and anti-drift

Task-board setup writes shared domain methodology for reviews and tests; each task adds its own goal, acceptance evidence, review/test strategy and bounded base work units. When a top-level task becomes active, the main session’s task-board flow announces and creates a unique hourly session anti-drift cron. Users can choose another cadence or opt out. Queued tasks keep their prepared methodology and complete task-specific prompt without a live timer.

Each delivered tick rereads the methodology, anti-drift rules and goal, collects active-agent updates, reconciles statuses and dependencies, and keeps all unfinished graph entries plus the latest 10 completed entries. Older completion evidence stays in task records. At most five short report lines give local time/timezone, tick number and elapsed time, achievements, remaining work and next action, blockers or questions when present, and the drift verdict. Completion, cancellation or parking stops the timer and verifies removal. Timers depend on the active session and runtime limits; unavailable scheduling is reported.

Bare +++ injects these planning steps only in Plan mode. File changes and scheduling wait until execution. See Task Board Setup and Manager Setup.

Agents

Artifact metadata

Every -setup skill stamps what it writes with the plugin version that generated it and generated_by naming the skill. JSON artifacts carry these plus last_updated as top-level keys. The byte-copied .mjs/.md hook and prompt files carry only version and generated_by in a // brewcode-meta: comment line (or .md frontmatter) — never last_updated, since the date would be the release date already present at the top of the file and rewriting it on every build would defeat the copy’s own drift check.

/brewcode:setup-status reads these stamps back across all eleven -setup skills in this marketplace and tells you when an install is running on an older brewtools version than the plugin currently installed — the fix is always that skill’s own upgrade mode.

Plugin variable

${CLAUDE_PLUGIN_ROOT} and ${CLAUDE_SKILL_DIR} are text substitutions Claude Code performs on the prompt itself, not environment variables — no hook injection required. Skills in the main conversation use ${CLAUDE_SKILL_DIR} to access their own files. Subagent .md definitions get ${CLAUDE_PLUGIN_ROOT} substituted into the prompt text at Agent spawn; it is not readable as a shell variable inside a Bash call.

Architecture

brewtools/
├── .claude-plugin/
│   └── plugin.json              # Plugin manifest
├── hooks/                       # filenames deliberately NOT renamed in v5.0.0
│   ├── hooks.json               # Hook registry
│   ├── session-start.mjs        # SessionStart: HARD-wall awareness, manager codeword prep
│   ├── manager-prompt.mjs       # Codewords; +++ only in Plan mode
│   ├── hardmode-guard.mjs       # Copied into a project by manager-setup install
│   └── lib/                     # utils.mjs, manager-state.mjs, manager-prompts.mjs
├── skills/
│   ├── text-optimize/           # Token optimization
│   ├── text-human/              # AI artifact removal
│   ├── secrets-scan/            # Secrets scanning
│   ├── ssh/                     # SSH server management
│   ├── deploy/                  # GitHub Actions deployment
│   ├── plugin-update/           # Check, install, and update plugin suite
│   ├── provider-switch/         # Alternative API provider management
│   ├── context-slim/            # Permanent-context compression
│   ├── agent-deadline-setup/    # Soft wall-clock budget for subagents
│   ├── agent-return-setup/      # Size budget on every subagent's final return
│   ├── agent-router-setup/      # Generic-spawn router (experimental)
│   ├── manager-setup/           # Hard delegation wall + codewords
│   └── task-board-setup/        # Kanban, methodology, graph and session anti-drift
└── agents/
    ├── text-optimizer.md        # Text optimization agent
    ├── ssh-admin.md             # SSH server administration
    └── deploy-admin.md          # GitHub Actions, CI/CD agent