mirror of
https://github.com/kochetkov-ma/claude-brewcode.git
synced 2026-09-14 20:16:41 +08:00
v5.0.0: -setup suffix on all setup skills, guide deleted, setup-status added, canonical mode vocabulary, critical-fix wave
This commit is contained in:
kochetkov-ma
+37
-11
@@ -4,8 +4,8 @@
|
||||
|
||||
| Field | Value |
|
||||
|-------|-------|
|
||||
| Version | 4.10.1 |
|
||||
| Skills | 8 |
|
||||
| Version | 5.0.0 |
|
||||
| Skills | 9 |
|
||||
| Agents | 5 |
|
||||
| Hooks | 2 |
|
||||
| Model | opus |
|
||||
@@ -45,7 +45,7 @@ Update anytime with `/brewtools:plugin-update`.
|
||||
|
||||
Brewcode turns single Claude Code sessions into an infinite task pipeline. Claude Code's native auto-compaction preserves the working context, and brewcode hooks re-inject plugin state on each session so the task runs to completion regardless of how many compaction cycles occur.
|
||||
|
||||
Skills cover semantic code search, multi-agent review, convention analysis, e2e orchestration, project rules, and meta-tooling for skills, agents and teams. The shipped agents are specialists only -- implementation, testing, review and architecture roles are generated per project by `/brewcode:teams`.
|
||||
Skills cover semantic code search, multi-agent review, convention analysis, e2e orchestration, project rules, and meta-tooling for skills, agents and teams. The shipped agents are specialists only -- implementation, testing, review and architecture roles are generated per project by `/brewcode:teams-setup`.
|
||||
|
||||
## Installation
|
||||
|
||||
@@ -65,23 +65,49 @@ claude --plugin-dir ./brewcode
|
||||
## Quick Start
|
||||
|
||||
```bash
|
||||
/brewcode:superreview # Generate a project-tailored deep-review skill
|
||||
/brewcode:setup-status # What is installed in this project, and what to run next
|
||||
/brewcode:superreview-setup # Generate a project-tailored deep-review skill
|
||||
```
|
||||
|
||||
## Skills
|
||||
|
||||
A `-setup` suffix marks a skill that **installs a mechanism you use afterwards instead of the skill** --
|
||||
`/brewcode:superreview-setup` emits a project-local `/superreview`, `/brewcode:teams-setup` writes agents you
|
||||
then delegate to. Recurring tools you invoke over and over (`agents`, `rules`, `skills`, `convention`, `e2e`)
|
||||
keep bare names.
|
||||
|
||||
Setup skills draw their modes from one vocabulary, in this order:
|
||||
|
||||
```
|
||||
status | install | upgrade | enable | disable | uninstall | purge
|
||||
```
|
||||
|
||||
No arguments = `status` when the mechanism is installed, `install` when it is not. The one exception is
|
||||
`/brewcode:semble-setup`, which **always** defaults to `status`, so a bare invocation can never start a
|
||||
machine-level package install.
|
||||
|
||||
Each skill implements the modes that mean something for it and rejects the rest with an error instead of
|
||||
guessing: `/brewcode:teams-setup` covers `status | install | upgrade | uninstall | purge` and rejects
|
||||
`enable` / `disable`, since a team either exists or it does not. Skill-specific extras come after the
|
||||
canonical set, never in place of it (`semble-setup`: `reindex | optimize | resume`).
|
||||
|
||||
| Skill | Purpose |
|
||||
|-------|---------|
|
||||
| [`/brewcode:superreview`](skills/superreview/README.md) | 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 |
|
||||
| [`/brewcode:teams`](skills/teams/README.md) | Dynamic agent team creation, management, and performance tracking -- every team also gets a fixed review-only `intent-guard` member (not counted in team size) |
|
||||
| [`/brewcode:setup-status`](skills/setup-status/README.md) | Read-only cross-plugin dashboard: which setup skills are installed, stale, partial or missing here, with the exact command to run for each. Runs no setup itself |
|
||||
| [`/brewcode:superreview-setup`](skills/superreview-setup/README.md) | 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 |
|
||||
| [`/brewcode:teams-setup`](skills/teams-setup/README.md) | Dynamic agent team creation, management, and performance tracking -- every team also gets a fixed review-only `intent-guard` member (not counted in team size) |
|
||||
| [`/brewcode:semble-setup`](skills/semble-setup/README.md) | Semantic code search setup: installs the pinned semble_code MCP, isolated cache, semble-first rule + hooks, agent migration |
|
||||
| [`/brewcode:convention`](skills/convention/README.md) | Extract etalon classes, patterns, architecture into convention docs and rules |
|
||||
| [`/brewcode:rules`](skills/rules/README.md) | Prompt-driven rules management: status, create, improve, review |
|
||||
| [`/brewcode:skills`](skills/skills/README.md) | Prompt-driven skill management: status, create, improve, sync, review |
|
||||
| [`/brewcode:agents`](skills/agents/README.md) | Prompt-driven agent management: status, create, improve, sync, review |
|
||||
| [`/brewcode:e2e`](skills/e2e/README.md) | E2E testing orchestration with BDD scenarios and quorum review |
|
||||
| [`/brewcode:semble`](skills/semble/README.md) | Semantic code search setup: installs the pinned semble_code MCP, isolated cache, semble-first rule + hooks, agent migration |
|
||||
| [`/brewcode:e2e`](skills/e2e/README.md) | E2E testing orchestration with BDD scenarios and quorum review. `install` writes the project's rules to `.claude/e2e/e2e-rules.md`; modes are `status \| install \| create \| update \| review \| rules` |
|
||||
|
||||
> **Note:** `/brewcode:superreview` emits a self-contained, project-local deep-review skill tailored to your stack.
|
||||
> **Run setups one at a time.** Each one is an interactive generator that fans out subagents and asks real
|
||||
> questions; two in a session degrade each other. `/brewcode:setup-status` tells you what is missing and prints
|
||||
> the command -- you run it yourself, ideally in a fresh session.
|
||||
|
||||
> **Note:** `/brewcode:superreview-setup` emits a self-contained, project-local deep-review skill tailored to your stack.
|
||||
> It always makes sure the project HAS domain experts (creating the missing ones via `agent-creator`), always emits
|
||||
> `.claude/agents/intent-guard.md`, then wires the emitted skill to the project's real gates, rules and scope
|
||||
> baseline (task + issue + recorded decisions). The EMITTED skill then resolves depth per run from your prompt:
|
||||
@@ -98,7 +124,7 @@ claude --plugin-dir ./brewcode
|
||||
| [bash-expert](agents/bash-expert.md) | inherit | Creates sh/bash scripts for Mac/Linux |
|
||||
| bc-rules-organizer | haiku | Internal: spawned by /brewcode:rules |
|
||||
|
||||
> **No generic agents:** brewcode ships specialists only. Implementation, testing, review and architecture work goes to project-specific agents in `.claude/agents/` — generate them with `/brewcode:teams create` (5-20 domain agents with self-selection protocol and performance tracking, plus one fixed review-only `intent-guard`).
|
||||
> **No generic agents:** brewcode ships specialists only. Implementation, testing, review and architecture work goes to project-specific agents in `.claude/agents/` — generate them with `/brewcode:teams-setup install` (5-20 domain agents with self-selection protocol and performance tracking, plus one fixed review-only `intent-guard`).
|
||||
|
||||
> **Scope guard:** every agent carries a `## Scope guard` -- if a task exceeds one bounded unit (one deliverable, ~5 files), the agent stops and proposes a split instead of running for an hour.
|
||||
|
||||
@@ -113,7 +139,7 @@ brewcode/
|
||||
| +-- hooks.json # Event bindings
|
||||
| +-- lib/utils.mjs # Shared utilities
|
||||
+-- agents/ # 5 agents
|
||||
+-- skills/ # 8 skills
|
||||
+-- skills/ # 9 skills
|
||||
+-- templates/ # Rule templates
|
||||
```
|
||||
|
||||
|
||||
Reference in New Issue
Block a user