Plan, delegate and review coding work through tiered Claude Code worker agents.
Multi-Agent & Subagents Skills for Claude Code, Codex & AI Agents
These skills help an agent split work, coordinate specialists, and collect their results. They include parallel-agent dispatch, harness building, councils, agent deployment, observability, and team roles.
20 skillsStart here
Create or revise concise repository-specific AGENTS.md instructions.
Claude-Mem Execute a Plan
Execute a phased plan with separate implementation, verification, quality review, and branch handoff responsibilities.
Design a project-specific harness with specialist agents, task Skills and a working orchestration structure.
Explore Multi-Agent & Subagents skills
16 skillsCollect execution feedback and improve an existing harness, its agents, Skills and orchestration rules.
software-factory
Use a shared vocabulary to design agent-operated software delivery systems.
Configure and troubleshoot tracing and logs for agents built with agents-cli.
Split independent problems into scoped agent tasks and integrate their results after review.
Execute a written implementation plan using task-scoped implementers and reviews with a progress ledger.
Compare perspectives from configured thinker personas to examine complex decisions and record a reasoned outcome.
Claude-Mem Branch Standup
Compare active worktrees, branches, and PRs in a shared read-only discussion, then produce one consolidation plan.
Plan and follow an agent lifecycle from requirements and scaffolding through evaluation and deployment.
Create or enhance agent projects with agents-cli scaffolds and optional deployment configuration.
Compare multiple agent-generated approaches to a problem.
Anti Slop — Core Anti Slop Rules
Apply the core antislop rules alongside focused design, writing and code-comment workflows.
chief-of-staff
Coordinate subagents and schedules around a long-running goal in one session.
Register an already deployed agent with Gemini and manage Agent Registry entries.
Write ADK Python or Go agents with patterns for tools, callbacks, state and orchestration.
Prepare agent deployment with agents-cli for Agent Runtime, Cloud Run or GKE.
Official workflows for building, evaluating, deploying and observing agents with Google Agents CLI.
Related topics
How to choose Multi-Agent & Subagents skills
Use dispatch and harness entries when work can run in parallel, council entries when viewpoints should be compared, and deployment entries when the team must run as a service.
- Split a project across parallel agents
- Build a harness for specialist agents
- Combine agent outputs into one decision


