subagent-delegation

• Leverage native parallel subagent dispatch and 200k+ context windows where available.

Activation Conditions

Use symptom -> action triggers: when one matches, apply this skill and verify with the protocol below.

Activate this skill when:

• Creating repetitive code structures or boilerplate
• Performing data transformation tasks
• Analyzing codebase for patterns or information
• Generating documentation from existing code
• Creating simple utility functions
• Breaking down complex features into manageable subtasks

Core Delegation Patterns

See Delegation Patterns for detailed examples of:

• Boilerplate generation (API routes, CRUD operations, component structures)
• Data transformations between formats
• File analysis and pattern extraction
• Documentation generation from code
• Utility function creation

<!-- MCP:START --> <!-- PORTABILITY:START -->

Cross-Client Portability

This skill is written to stay usable across GitHub Copilot, Claude Code, and Codex.

• GitHub Copilot: keep the folder in a Copilot-visible skill path or wrap the workflow in project instructions when folder discovery is unavailable.
• Claude Code: keep the folder in a local skills directory or a compatible plugin source.
• Codex: install or sync the folder into $CODEX_HOME/skills/subagent-delegation and restart Codex after major changes.

<!-- PORTABILITY:END -->

MCP Availability And Fallback

Preferred MCP Server: None required

• Fallback prompt: "Use the Subagent Delegation Patterns skill without MCP. Rely on its local instructions, bundled resources, standard shell or editor tools, and direct verification. Show the evidence used before concluding."
• Do not claim an MCP operation was used when the active host does not expose it.
• Treat local files, tests, rendered outputs, logs, or screenshots as the fallback evidence path.

<!-- MCP:END -->

Anti-Patterns

• Delegating or evaluating without a scoped success condition: The output becomes hard to review and easy to overbuild.
• Skipping the evidence step: A workflow that cannot be re-checked quickly is not ready for handoff.
• Bundling unrelated subtasks together: It creates noisy prompts, weaker ownership, and avoidable integration risk.

Verification Protocol

Before claiming "skill applied successfully":

1. Pass/fail: The Subagent Delegation workflow names the agent boundary, delegated scope, and expected return artifact.
2. Pass/fail: Context passed to helpers is minimal, task-local, and free of hidden expected answers.
3. Pass/fail: Results are integrated only after evidence, diffs, or citations are checked by the controller.
4. Pressure-test scenario: Run the workflow on two similar tasks that must not share assumptions or leaked context.
5. Success metric: Zero context leakage; every delegated output is independently reviewable.

Examples & Scripts

• Delegation Pattern Examples — Code examples of common delegation patterns
• Delegation Template — JavaScript template for structuring delegation calls

Self-Verification Phase-Gate Questions

Before you mark delegated work complete, the coordinating agent must ask:

• Did the subagent receive a scoped task, owned files or responsibilities, and a clear definition of done?
• Did I review the returned output for correctness, integration fit, and missing edge cases instead of forwarding it blindly?
• Did I preserve the evidence trail showing what was delegated, what changed, and what still needs local ownership?
• Are any follow-up fixes, risks, or unresolved assumptions now explicit to the next reviewer?

Integration Workflow

For delegating tasks to subagents, follow this 5-step process:

• Step 1: Plan - Analyze problem, design solution, identify routine subtasks
• Step 2: Delegate - Use runSubagent for routine or repetitive work
• Step 3: Review - Check output for correctness, completeness, integration compatibility
• Step 4: Integrate - Incorporate output into codebase, handle conflicts
• Step 5: Validate - Test integrated code, debug issues, ensure quality

Quality Control

Before integrating subagent results, verify:

• [ ] Code follows project conventions (style, naming, structure)
• [ ] Matches specified interfaces and contracts
• [ ] Includes necessary error handling
• [ ] Has appropriate comments/documentation
• [ ] No security vulnerabilities or obvious performance issues
• [ ] Compatible with existing codebase (imports, dependencies)

Combining with Sequential Thinking

For complex tasks, use Sequential Thinking first to plan architecture, then delegate routine parts:

// Use Sequential Thinking to plan
mcp_sequentialthi_sequentialthinking({
  thought: "Breaking down feature...identifying repetitive CRUD for delegation",
  thoughtNumber: 1,
  totalThoughts: 5,
  nextThoughtNeeded: true
})

// Then delegate boilerplate
runSubagent({
  description: "Generate CRUD API",
  prompt: "Create CRUD functions matching data model..."
})

// Main agent implements core logic

---

Related Skills

• agent-task-mapping: Use it when the workflow also needs task-to-agent routing decisions.
• custom-agent-usage: Use it when the workflow also needs loading and invoking custom agent definitions safely.
• subagent-driven-development: Use it when the workflow also needs plan-driven implementation with reviewer loops.
• agentic-eval: Use it when the workflow also needs rubric-driven evaluation loops.