skills.rest
bennybennison
bennybennisonCOMMUNITY
Explore all skills by this author →
GitHub

agent-orchestration

Orchestrate specialist agent stages with structured handoff documents and evidence bundles.

Updated Apr 11, 2026
Search Tags:#code review#agents#cleanup#orchestration#pipeline#handoff
One-click install
npx skills add https://github.com/bennybennison/agent-toolkit --skill agent-orchestration-bennybennison
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-orchestration
Source: https://github.com/bennybennison/agent-toolkit/tree/main/skills/agent-orchestration
Command: npx skills add https://github.com/bennybennison/agent-toolkit --skill agent-orchestration-bennybennison

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides repeatable patterns to break multi-step software work into specialist agent stages and ensure clear, evidence-backed handoffs so work proceeds without author-bias, lost context, or unchecked failures.

Core Features & Use Cases

  • Pipeline patterns: Defines feature, bug-fix, and refactor pipelines with explicit stage roles (planner/implementer/cleanup/reviewer).
  • Structured handoffs: Offers a handoff template and guidance for bundling file paths, diffs, tests, decisions, and open questions so the next agent can act deterministically.
  • Operational principles: Enforces author-bias elimination, fail-fast stopping, minimal scope per agent, and a cleanup pass before review.
  • Use Case: Use to coordinate planning, implementation, cleanup, and a separate code review when delivering a feature or regression fix.

Quick Start

Plan and run a feature pipeline: have an audit-planner produce acceptance criteria and risks, invoke an implementer to apply changes, run a cleanup pass, then hand off a structured document and evidence bundle to a separate code-reviewer.

Frequently Asked Questions about agent-orchestration

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

FAQPage Schema
How do I orchestrate multiple agents for a multi-step software feature implementation?▼

To orchestrate multi-step software feature implementation, you chain specialist agents into structured workflows with explicit handoff documents separating planning, implementation, cleanup, and review stages. This ensures deterministic progression without author-bias or lost context.

What is the best way to structure handoffs between an implementer agent and a code reviewer?▼

The best way to structure handoffs for code review is using a structured document bundling file paths, diffs, tests, decisions, and open questions. This separates author and reviewer contexts, ensuring the reviewing agent acts deterministically on evidence.

How do agent pipelines handle stage failures during a bug fix or refactor?▼

Agent pipelines handle stage failures during a bug fix or refactor using fail-fast behavior. If any stage encounters a failure, the pipeline stops immediately to prevent unchecked errors from propagating to downstream agents.

Do I need a separate cleanup pass before running an automated code review?▼

Yes, you need a separate cleanup pass before running an automated code review. Enforcing a cleanup stage ensures the codebase is normalized and minimal scope is maintained before the reviewer agent evaluates the implementation evidence.

What should be included in an evidence bundle for an agent code review handoff?▼

An evidence bundle for an agent code review handoff should include specific file paths, code diffs, executed tests, architectural decisions, and any remaining open questions. This provides the reviewer agent with the exact context needed for validation.

Related Skills

Explore More Software Engineering Skills→
tomguntomgun

planning-features

Create implementation plans for software features with scope, files, acceptance criteria, and risks.

Community
Intermediate
lukasdeveloperacclukasdeveloperacc

code-reviewer

Review code diffs and implementation evidence in Korean via Codex.

Community
Intermediate
snitssnits
1

domain-review-before-implementation

Review task briefs for domain-specific issues before code implementation.

Community
Basic
rlagycksrlagycks
3

claude-devfleet

Orchestrate parallel AI coding agents across structured software development missions.

Community
Advanced