orchestrating-swarms

Coordinate leaderless cross-repo investigations and parallel swarms via pre-launch contracts and workspace memos.

Updated Apr 23, 2026
One-click install
npx skills add https://github.com/AndurilCode/swarmspace --skill orchestrating-swarms-andurilcode
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: orchestrating-swarms
Source: https://github.com/AndurilCode/swarmspace/tree/main/skills/orchestrating-swarms
Command: npx skills add https://github.com/AndurilCode/swarmspace --skill orchestrating-swarms-andurilcode

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinate cross-repo investigations and parallel swarms by configuring a pre-launch contract and memo-driven communication. This enables orchestration of multi-repo tasks without a centralized leader, letting peers operate as equals and synthesize results after completion.

Core Features & Use Cases

  • Leaderless coordination via a filesystem memo + journal, enabling parallel exploration and risky changes across multiple repos.
  • Supports three primitives: ws investigate, ws swarm, and ws agent, each with its own protocol and output patterns.
  • Pre-launch contract: define peer count, backend CLI, and post-run synthesis to ensure consistent expectations before execution.
  • Post-run synthesis: results are written into the workspace memo (summary, audit-result, finding-*) for easy review and auditing.
  • Ideal for cross-service bug diagnosis, multi-repo refactors or migrations, and any scenario requiring scoped, parallel work across many repos.

Quick Start

Boot a workspace, specify N, and launch ws investigate or ws swarm to begin a parallel, leaderless cross-repo effort.

Frequently Asked Questions about orchestrating-swarms

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

FAQPage Schema
How do I coordinate parallel investigations across multiple repositories?

You can coordinate parallel investigations across multiple repositories by configuring a pre-launch contract and using filesystem memo-driven communication, letting peers operate as equals without a centralized leader.

What is the best way to diagnose cross-service bugs spanning multiple repositories?

Diagnosing cross-service bugs is best handled by launching parallel swarms that explore scoped code paths across multiple services, recording findings into a workspace memo for post-run synthesis and auditing.

How do I configure a pre-launch contract for a multi-repo swarm?

Configure a pre-launch contract by defining the peer count, backend CLI, and post-run synthesis expectations before execution, ensuring consistent workspace behavior across all parallel agents.

Can I audit deprecated APIs across a workspace with multiple peers?

Yes, you can audit deprecated APIs across a workspace by launching parallel agents that investigate each service independently, writing audit results into a shared workspace memo for easy review.

How does leaderless swarm orchestration work for workspace tasks?

Leaderless swarm orchestration works by having peers communicate through a filesystem memo and journal, enabling parallel exploration and risky changes across multiple repos without a centralized leader coordinating the effort.

When should I use ws investigate versus ws swarm for cross-repo tasks?

Use ws investigate, ws swarm, or ws agent depending on your cross-repo task protocol, each supporting parallel multi-repo work with distinct output patterns recorded in the workspace memo for synthesis.