Files
kochetkov-ma__claude-brewcode/brewcode
..
2026-03-05 17:29:18 +00:00

Brewcode

Infinite task execution plugin for Claude Code -- automatic context handoff, multi-agent workflows, knowledge persistence.

Field Value
Version 3.14.0
Skills 15
Agents 12+
Hooks 8
Model opus

Install

Paste this into a Claude Code session:

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

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

After install, run `/reload-plugins` (or `exit` + `claude`).
Or install the whole 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.

Update anytime with /brewtools:plugin-update.

Overview

Brewcode turns single Claude Code sessions into an infinite task pipeline. When context reaches ~90%, the PreCompact hook saves knowledge, writes handoff state, and the session continues automatically. One cycle: spec -- plan -- start -- and the task runs to completion regardless of how many compaction cycles occur.

14 skills cover the full lifecycle: project analysis, specification creation through parallel research agents, phased plan generation with quorum review, execution with automatic handoff, code review, convention analysis, and project rules management. 12+ specialized agents handle implementation, testing, review, architecture, and coordination.

Installation

# Marketplace (recommended)
claude plugin marketplace add https://github.com/kochetkov-ma/claude-brewcode
claude plugin install brewcode@claude-brewcode

# Already installed? Update
claude plugin marketplace update claude-brewcode
claude plugin update brewcode@claude-brewcode

# Dev mode (no install)
claude --plugin-dir ./brewcode

Quick Start

/brewcode:setup                              # 1. Adapt templates for the project (one-time)
/brewcode:spec "Implement JWT authorization"  # 2. Research + specification
/brewcode:plan                                # 3. Generate phased plan
/brewcode:start                               # 4. Execute with infinite context

After /brewcode:setup, each task follows the cycle: spec -> plan -> start.

Skills

Skill Purpose
/brewcode:setup Analyze project, check prerequisites, generate adapted templates and config
/brewcode:spec Research codebase + user dialog -> SPEC.md
/brewcode:plan Generate phased PLAN.md from SPEC or Plan Mode with quorum review
/brewcode:start Execute task with infinite context through automatic handoffs
/brewcode:teams Dynamic agent team creation, management, and performance tracking
/brewcode:standards-review Review code for project standards compliance
/brewcode:convention Extract etalon classes, patterns, architecture into convention docs and rules
/brewcode:rules Extract rules from accumulated knowledge to .claude/rules/
/brewcode:grepai Semantic code search (setup, status, start, stop, reindex)
/brewcode:skills Skill management: list, create, upgrade with activation optimization
/brewcode:agents Interactive agent creation and improvement
/brewcode:e2e E2E testing orchestration with BDD scenarios and quorum review
/brewcode:glm-design-to-code Vision model design-to-code: screenshot/text/HTML/URL to multi-framework code
/brewcode:debate Evidence-based multi-agent debate: Challenge, Strategy, Critic modes
/brewcode:teardown Plugin configuration cleanup (tasks are preserved)

Note: /brewcode:review is a local skill created in the project during /brewcode:setup.

Agents

Agent Model Purpose
developer opus Implement features, write code, fix bugs
tester sonnet Run tests, analyze failures, debug flaky tests
reviewer opus Code review, architecture, security, performance
architect opus Architecture analysis, patterns, trade-offs, scaling
skill-creator opus Create and improve Claude Code skills
agent-creator opus Create and improve Claude Code agents
hook-creator opus Create and debug Claude Code hooks
bash-expert opus Create professional shell scripts
bc-coordinator haiku Task coordination, artifact management
bc-knowledge-manager haiku KNOWLEDGE.jsonl compaction and deduplication
bc-grepai-configurator opus Generate grepai config.yaml
bc-rules-organizer haiku Create and optimize .claude/rules/ files

Dynamic teams: Use /brewcode:teams create to generate 5-20 project-specific agents with self-selection protocol and performance tracking.

Architecture

brewcode/
+-- .claude-plugin/plugin.json          # Plugin manifest
+-- hooks/                              # 8 lifecycle hooks
|   +-- session-start.mjs              # Session initialization
|   +-- grepai-session.mjs             # Auto-start grepai watch
|   +-- pre-task.mjs                   # Knowledge injection into agents
|   +-- grepai-reminder.mjs            # grepai reminder
|   +-- post-task.mjs                  # Session binding, 2-step protocol
|   +-- pre-compact.mjs               # Knowledge compaction, handoff
|   +-- stop.mjs                       # Exit blocking
|   +-- forced-eval.mjs                # Skill activation
+-- agents/                            # 12 agents
+-- skills/                            # 15 skills
+-- templates/                         # Rule templates

Hook Lifecycle

Hook Event Purpose
session-start SessionStart Initialize session, inject plugin path
grepai-session SessionStart Auto-start grepai watch process
pre-task PreToolUse:Task Inject grepai + KNOWLEDGE into agent prompts
grepai-reminder PreToolUse:Glob/Grep Remind to prefer semantic search
post-task PostToolUse:Task Bind session, enforce 2-step protocol
pre-compact PreCompact Compact KNOWLEDGE, write handoff entry
stop Stop Block if not terminal, clean lock
forced-eval UserPromptSubmit Skill activation

Task Structure

.claude/tasks/{TS}_{NAME}_task/
  SPEC.md             # Specification (research results)
  PLAN.md             # Phased execution plan
  KNOWLEDGE.jsonl     # Accumulated knowledge (survives compactions)
  phases/             # Individual phase files for agents
  artifacts/          # Execution reports by phases
  backup/             # Backups
  .lock               # Session lock file

Documentation

Full docs: doc-claude.brewcode.app/brewcode/overview

Resource Link
Skills reference Skills
Agents reference Agents
Hooks reference Hooks
Workflow Workflow
Release Notes RELEASE-NOTES.md

Author: Maksim Kochetkov | License: MIT