# Coordinate Parallel Subagent Workflows

Multi-agent work can fail when tasks overlap or depend on hidden context. This skill helps Claude, Codex, and Claude Code plan bounded subagent assignments.

## Install

```bash
npx skillstore add diegopherlt/subagent-orchestration
```

## Metadata

- Status: approved
- Slug: diegopherlt-subagent-orchestration
- Skillstore revision: r1
- Version status: missing
- Tree hash: 0cbb8c74e4728bf5fbe06876eec0130cce6e128cf12ad4f8eec2074bd087de99
- Author: DieGopherLT
- GitHub username: DieGopherLT
- License: MIT
- Repository: https://github.com/DieGopherLT/claude-kit/tree/main/skills/subagent-orchestration
- Ref: 72d5025b022c77f7a51bdf5c1637c689c80e89d1
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 79
- Quality tier: bronze
- Public page: https://skillstore.pages.dev/skills/diegopherlt-subagent-orchestration
- Manifest: https://skillstore.pages.dev/api/skills/diegopherlt-subagent-orchestration/manifest

## Capabilities

- Defines when independent subagent tasks can run in parallel.
- Provides a reusable scope template for delegated work.
- Encourages explicit file lists and exclusion lists.
- Separates parallel work from sequential work with dependency checks.
- Guides users to include reference documents for subagents.

## Use Cases

- Split independent code reviews: Assign separate subagents to review unrelated modules with clear files to inspect and exclude.
- Plan parallel implementation tasks: Define non-overlapping feature work before asking agents to edit different areas of a repository.
- Control delegated research: Give each research agent a bounded question, required references, and clear limits on what to ignore.

## Prompt Templates

### Identify safe delegation

```
Review this task and tell me whether it should use one agent or multiple subagents. Explain any dependencies and scope risks.
```

### Create bounded subagent tasks

```
Break this work into independent subagent tasks. For each task, list files to modify, files not to touch, constraints, and reference documents.
```

### Validate parallel readiness

```
Evaluate these proposed subagent assignments for overlap, missing exclusions, shared dependencies, and unclear ownership. Revise them for safe parallel execution.
```

### Design an orchestration plan

```
Create a multi-stage orchestration plan for this project. Separate parallel phases from sequential phases and define merge checks after each phase.
```

## Limitations

- Does not launch or manage subagents by itself.
- Does not inspect repositories or verify task independence automatically.
- Requires the user to provide accurate file paths and boundaries.
- Cannot resolve conflicts after subagents modify the same files.

## Best Practices

- Use parallel subagents only when tasks are independent and bounded.
- Name exact files to modify and exact files to avoid.
- Include reference documents so each subagent has enough context.

## Anti Patterns

- Do not assign vague work such as updating related files.
- Do not run dependent tasks in parallel.
- Do not omit exclusions when multiple agents work in one repository.

## Security Audit

- Audited at: 2026-07-06T11:40:03.473\+00:00
- Summary: All three static findings are false positives caused by documentation language, code fences, and the phrase 'Files NOT to touch.' The skill provides workflow guidance only and does not execute commands, access networks, or request secrets. No prompt injection or malicious intent was found in the reviewed files.

## Stats

- Views: 176
- Downloads: 10
- Favorites: 0
- Popularity score: 0