Destynova2/cli-code-skills / cli-forge-chef
cli-forge-chef
VerifiedGenerate a multi-agent orchestrator for any project using the Brigade de Cuisine pattern: Chef (plans, decides), Sous-Chef (validates, merges), Commis (execute). Clients order via a Menu (PERT), the kitchen delivers Plats (merged features). Creates tmuxinator config, prompts, shared-state memory, quality gates, and permissions. Uses Claude Code Agent Teams (TeamCreate + SendMessage) with visible tmux panes. Use when the user wants to parallelize work across multiple Claude Code agents, orchestrate a multi-agent sprint, or says 'boss', 'multi-agent', 'orchestrate', 'parallelize agents', 'team of claudes', 'sprint multi-agent', 'brigade', 'chef'. Also triggers on 'swarm', 'agent team', 'parallel workers', 'PERT multi-agent'.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 51.4 KB |
| references/anti-patterns-chef.md | markdown | 2.7 KB |
| references/apply-quorum.md | markdown | 9.3 KB |
| references/brigade-setup-worktrees.shexecutable | script | 4.7 KB |
| references/ccheck-prompt-template.md | markdown | 4.2 KB |
| references/chef-prompt-template.md | markdown | 37.6 KB |
| references/conflict-resolution.md | markdown | 10.1 KB |
| references/contre-chef-inter-prompt-template.md | markdown | 8.9 KB |
| references/gotchas-chef.md | markdown | 38.0 KB |
| references/maitre-dhotel.md | markdown | 14.4 KB |
| references/parallel-exploration.md | markdown | 9.1 KB |
| references/permissions-template.md | markdown | 16.7 KB |
| references/pert-computation.md | markdown | 17.5 KB |
| references/quality-gates.md | markdown | 1.9 KB |
| references/raw-tmux-fallback.md | markdown | 3.3 KB |
| references/shared-state-template.md | markdown | 8.4 KB |
| references/simplified-model.md | markdown | 9.8 KB |
| references/sprint-persistence.md | markdown | 7.0 KB |
| references/sprint-report-template.md | markdown | 8.0 KB |
| references/templates.md | markdown | 762 B |
| references/tmuxinator-template.md | markdown | 7.4 KB |
references/anti-patterns-chef.md
View on GitHub# Brigade Anti-Patterns — Detection and Resolution
> **When to read:** During execution (Phase 2+), to identify and fix brigade dysfunctions.
---
## Named anti-patterns
| # | Name | Detection | Severity | Resolution |
|---|------|-----------|----------|------------|
| B1 | **Ping-Pong Rejection** | Same commis × same sous-chef → rejects the same diff 3+ times | Critical | Patch Bankruptcy: reassign or redefine the spec |
| B2 | **Hot File Thrashing** | 3+ commis edit the same file during the same sprint | Major | Sequence the merges (G15); mark the file as a sensitive zone (3/3) |
| B3 | **Ghost Commis** | Commis assigned but no commit/message for > 30min | Major | Timeout, reassign, diagnose (NFD: can it communicate?) |
| B4 | **Gate Consensus Flip** | Sous-chef approves a diff then rejects it in the next round (or vice versa) | Major | Audit the sous-chef — its rules are inconsistent |
| B5 | **God Commis** | Commis touches > 50% of sprint files OR > 5 independent tasks | Major | Split, reassign, add another commis |
| B6 | **Speculative Task** | Commis starts before its dependency tokens are dropped | Minor | Remind the stigmergy rule; if intentional, mark as EXPLORATION |
| B7 | **Dead Branch** | A commis's branch has existed > 2 sprints without merge | Minor | Archive, clean up, notify |
| B8 | **Inter-Commis Feature Envy** | Commis A depends on > 3 outputs from Commis B | Major | Co-assign to the same commis OR precompute the shared dependency |
| B9 | **Sous-Chef Bypass** | Commis merges directly without going through the vote | Critical | Block the permissions; a commis must NEVER push to the main branch |
| B10 | **Over-Staffing** | 5+ commis for < 3 independent tasks | Minor | Shrink the brigade (mitosis tier S/M) |
---
## Detection heuristics
### B1 — Ping-Pong Rejection
```
For each (commis, sous-chef) pair:
count = number of consecutive rejections on the same file/diff
IF count >= 3 → PATCH BANKRUPTCY
```
### B3 — Ghost Commis
```
For each active commis:
last_activity = max(last commit, last SendMessage, last shared-state edit)
IF now - last_activity > 30min → GHOST
NFD diagnosis:
1. Can the commis read shared-state.md? (read test)
2. Can the commis write to its worktree? (write test)
3. Can the commis send a SendMessage? (comm test)
IF any fails → communication problem, not performance
```
### B5 — God Commis
```
For each commis:
files_ratio = files_touched(commis) / total_files_sprint
task_count = number of assigned tasks
IF files_ratio > 0.5 OR task_count > 5 → GOD COMMIS
```
### B8 — Inter-Commis Feature Envy
```
For each (commis_A, commis_B) pair:
deps = number of tokens from A consumed by B
IF deps > 3 → FEATURE ENVY
→ Co-assign or precompute
```
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 468k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |