fleet

Coordinate multiple AI agents across isolated workspaces in parallel waves.

820|78|Updated Mar 20, 2026
One-click install
npx skills add https://github.com/SethGammon/Citadel --skill fleet-sethgammon
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: fleet
Source: https://github.com/SethGammon/Citadel/tree/main/skills/fleet
Command: npx skills add https://github.com/SethGammon/Citadel --skill fleet-sethgammon

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Citadel Fleet coordinates parallel campaigns via waves, orchestrating multiple AI agents and preserving cross-wave memory to optimize throughput.

Core Features & Use Cases

  • Parallel waves orchestration: spawn 2-3 agents per wave in isolated worktrees
  • Momentum/memory: cross-session discovery relay to prevent rework
  • Modes: standard, quick, and speculative modes for flexible workflows

Quick Start

Launch a Fleet session to orchestrate parallel campaigns across waves.

Frequently Asked Questions about fleet

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

FAQPage Schema
How do I coordinate parallel AI agents to run concurrent campaigns?

You can orchestrate multiple AI agents to run parallel campaigns in isolated workspaces, enforcing scope non-overlap within each wave and sharing discoveries across sessions to preserve momentum.

What is wave management for parallel execution?

Wave management splits large projects into independent streams, spawning 2-3 agents per wave in isolated worktrees to execute concurrently, compressing discoveries and logging telemetry between waves.

Can I run speculative and quick modes with configurable timeouts for agent orchestration?

Yes, agent orchestration supports standard, quick, and speculative modes with configurable timeouts, allowing you to adjust parallel execution workflows based on your project's velocity requirements.

How does session memory prevent rework when splitting projects into independent streams?

Session memory generates session files and compresses discoveries to relay cross-wave findings, preventing rework by ensuring completed wave insights are shared with subsequent independent streams.

Does parallel execution work without external dependencies for isolated workspaces?

Yes, the parallel execution orchestration operates with no external dependencies, directly managing isolated workspaces, telemetry logging, and scope enforcement within its core framework.