orchestrate

Coordinate a six-agent pipeline with feedback loops and quality gates.

1|Updated Mar 15, 2026
One-click install
npx skills add https://github.com/kegbi/monica-companion --skill orchestrate-kegbi
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: orchestrate
Source: https://github.com/kegbi/monica-companion/tree/main/.claude/skills/orchestrate
Command: npx skills add https://github.com/kegbi/monica-companion --skill orchestrate-kegbi

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Orchestrate coordinates a six-agent pipeline to execute roadmap tasks with feedback loops and quality gates.

Core Features & Use Cases

  • Multi-agent pipeline: Planner, Plan Reviewer, Implementer, Code Reviewer, Smoke Tester, Committer, each with clean context and file-based communication.
  • Stepwise governance: Enforces planning → implementing → reviewing → testing → committing with retries and hard success criteria.
  • Robust state management: Tracks per-task progress in .claude-work/{task-id}, supports manual and auto modes, and resumes where left off.

Quick Start

Run /orchestrate with the desired mode to begin processing roadmap tasks from the first unchecked group.

Frequently Asked Questions about orchestrate

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

FAQPage Schema
How do I automate a multi-agent workflow to execute complex roadmaps end-to-end?

You can automate complex roadmap execution by running a six-agent pipeline that coordinates planning, implementing, code review, smoke testing, and committing stages with strict sequencing and quality gates. Use the orchestrate command to begin processing tasks from the first unchecked group.

What is a multi-agent pipeline for software development roadmaps?

A multi-agent pipeline is a coordinated workflow using specialized agents like Planner, Plan Reviewer, Implementer, Code Reviewer, Smoke Tester, and Committer. Each agent operates with clean context and file-based communication to execute roadmap tasks step by step.

How do I resume a paused roadmap pipeline task if my workflow is interrupted?

The pipeline resumes where it left off by tracking per-task progress in the .claude-work/{task-id} directory. It stores state locally, allowing the workflow to recover automatically and continue executing the remaining stages without losing completed work.

Can I use manual mode to control stepwise governance in an automated workflow?

Yes, the pipeline supports both manual and auto modes. In manual mode, you control the progression through planning, implementing, reviewing, testing, and committing stages, enforcing hard success criteria and retries before moving to the next step.

What's the best way to enforce quality gates and code reviews in an automated pipeline?

To enforce quality gates, use a stepwise pipeline that separates code review and smoke testing into distinct stages. The workflow enforces hard success criteria, applies retry rules, and audits artifacts before allowing a commit.

Why does my multi-agent pipeline fail to move past the planning stage?

The pipeline enforces strict sequencing and hard success criteria. If the Plan Reviewer stage fails, the workflow applies retry rules before proceeding to the implementing stage, ensuring artifact auditing and quality gates are met.