delegate-work-to-subagents

Orchestrates multi-agent workflows by delegating research, implementation, QA, and verification to subagents.

3|Updated Feb 15, 2025
One-click install
npx skills add https://github.com/mark-torres10/ai_tools --skill delegate-work-to-subagents-mark-torres10
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: delegate-work-to-subagents
Source: https://github.com/mark-torres10/ai_tools/tree/main/skills/delegate-work-to-subagents
Command: npx skills add https://github.com/mark-torres10/ai_tools --skill delegate-work-to-subagents-mark-torres10

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Coordinating complex, multi-part engineering work in a single agent session leads to context overload, file conflicts, and missed verification steps. This Skill turns the current agent into an orchestrator that splits work into discrete tasks and hands each one to a dedicated Composer 2.5 subagent. ## Core Features & Use Cases - Structured Task Decomposition: Splits work into research, implementation, QA, verification, integration check, and status reporting subagents, each with a defined role and constraints. - File Ownership Enforcement: Assigns each implementation subagent a closed set of files it may edit, preventing parallel subagents from colliding on the same files. - Checkpointed Status Reporting: Spawns status synthesis subagents after research and implementation waves so the user receives concise progress updates instead of raw logs. - Use Case: When implementing a feature that touches multiple modules, the orchestrator spawns a research subagent to map the codebase, parallel implementation subagents with non-overlapping file sets, an integration check, then QA and verification subagents before reporting completion. ## Quick Start Ask the agent to delegate this multi-part change to subagents and orchestrate the research, implementation, QA, and verification work.

Frequently Asked Questions about delegate-work-to-subagents

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I delegate a multi-task code change to subagents?

Activate this Skill and describe the change. The orchestrator builds a checklist, assigns each task to a research, implementation, QA, or verification subagent, and runs independent tasks in parallel while tracking completion.

How does subagent orchestration prevent file conflicts between parallel tasks?

Each implementation subagent receives a closed set of files it may edit. Tasks that would touch the same file run in sequence instead of parallel, and the orchestrator verifies after each wave that no subagent edited outside its set.

Which model do the subagents use in this orchestration workflow?

Every subagent uses Composer 2.5 (model: composer-2.5). Only the orchestrator may spawn subagents, and a subagent is not allowed to spawn another subagent.

What does each subagent return after finishing its task?

Each subagent returns an executive summary covering what it did, files or sources touched, tests run and results, remaining blockers, risks, and suggested next steps. Full diffs, log dumps, and long transcripts are excluded.

When should I not use subagent delegation for a coding task?

Avoid it for small single-file edits or tasks that cannot be cleanly split, since orchestration overhead outweighs the benefit. It is designed for work that divides into research, implementation, QA, and verification units.