Skip to content
Back to skills

Team Config

ASecurity

Set up parallel agent team for COMPLEX tasks. Creates worktree structure, generates ownership map, and initializes task queue. Use when orchestrator decomposes a task into parallel subtasks, or when asked to "set up team", "parallel setup", or "configure worktrees".

  • 14 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added May 29, 2026
ai-agentsgoshellbashgit

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned May 29, 2026

npx -y skills add ngocsangyem/MeowKit --skill team-config --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Team Config?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Team Config
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/ngocsangyem-team-config/badge)](https://www.skillsdirectory.com/skills/ngocsangyem-team-config)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: mk:team-config
description: >-
  Set up parallel agent team for COMPLEX tasks. Creates worktree structure,
  generates ownership map, and initializes task queue. Use when orchestrator
  decomposes a task into parallel subtasks, or when asked to "set up team",
  "parallel setup", or "configure worktrees".
argument-hint: '"task description" --agents N'
source: local
allowed-tools:
  - Read
  - Glob
  - Bash
  - Write
keywords: [team-config, worktree-setup, ownership-map, parallel-team-init, task-queue-init]
when_to_use: "Auto-invoked when orchestrator decomposes a COMPLEX task into parallel subtasks — sets up worktrees, ownership map, task queue. Not user-callable directly."
user-invocable: false
---

# Team Configuration

Sets up the infrastructure for parallel agent execution. Called by orchestrator when a COMPLEX task is decomposed into independent subtasks.

## What It Does

1. **Analyzes the task** — identifies independent subtasks from the plan
2. **Generates ownership map** — assigns file ownership globs per subtask
3. **Creates worktrees** — one git worktree per parallel agent
4. **Initializes task queue** — `session-state/task-queue.json` with claiming protocol
5. **Validates no overlap** — checks ownership globs for conflicts before starting

## Setup Flow

```
Plan approved (Gate 1)
    ↓
Orchestrator identifies parallel opportunity
    ↓
mk:team-config
    ├── 1. Parse plan for independent subtasks
    ├── 2. Generate ownership map (which agent owns which files)
    ├── 3. Validate zero overlap between ownerships
    ├── 4. Create git worktrees via mk:worktree
    └── 5. Create task-queue.json via mk:task-queue
    ↓
Parallel agents start claiming tasks
```

## Ownership Map Template

See `templates/ownership-map-template.md` for the format agents use to declare file ownership.

## Constraints

- Max 3 parallel agents (from `parallel-execution-rules.md`)
- Only COMPLEX tasks qualify (orchestrator enforces this)
- Gates (1 and 2) are never parallelized
- All worktrees branch from current feature branch HEAD

## Gotchas

- **Overlapping ownership kills parallelism** — if two subtasks need the same file, they can't be parallel. Restructure the decomposition or handle the shared file in a sequential pre-step
- **Worktree naming with special characters** — skill names with `:` (e.g., `mk:review`) need quoting in shell paths
- **Stale worktrees from crashed sessions** — run `git worktree list` to check for orphaned worktrees before creating new ones
- **Integration test is mandatory** — after merging all worktrees, the full test suite MUST pass. Don't skip this

## Teardown

After all parallel agents complete and integration test passes:
1. Merge all worktree branches to feature branch
2. Remove worktrees via `mk:worktree cleanup`
3. Archive `session-state/task-queue.json`

Files in this skill

  • SKILL.md2.8 KB
  • templates/ownership-map-template.md832 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.