
Fleet
- 1 repo stars
- Updated August 2, 2026
- g6xai/claude-plugins
Software engineering operations platform. Bootstrap any repo for autonomous development with process enforcement, Linear/Notion sync, and parallel TDD execution.
About
fleet is a Claude Code skill in the AI & Agent Building category. Software engineering operations platform. Bootstrap any repo for autonomous development with process enforcement, Linear/Notion sync, and parallel TDD execution.
- fleet
- AI & Agent Building
- AI-coding skill
Fleet by the numbers
- Data as of Aug 3, 2026 (Skillselion catalog sync)
/plugin marketplace add g6xai/claude-plugins/plugin install fleet@g6xaiAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| repo stars | ★ 1 |
|---|---|
| Last updated | August 2, 2026 |
| Repository | g6xai/claude-plugins ↗ |
What it does
Software engineering operations platform. Bootstrap any repo for autonomous development with process enforcement, Linear/Notion sync, and parallel TDD execution.
README.md
Fleet
Software Engineering Operations Platform for Claude Code
Fleet bootstraps any repository — greenfield or brownfield — for fully autonomous, process-compliant development. One plugin, works everywhere.
Install
See the parent repo README for install instructions.
What Fleet Does
- Greenfield repos: Detects BMAD, guides you through product brief → PRD → architecture → epics/stories
- Brownfield repos: Discovers tech stack, audits what's real vs stub, generates specs from gaps
- Both: Sets up test infrastructure, CI gates, process enforcement hooks, then runs parallel TDD agents
- Always: Syncs to Linear (tasks) and Notion (docs) with repo as single source of truth
Relationship to BMAD
Fleet works alongside BMAD — it does not install, modify, or bundle BMAD. BMAD is a separate upstream framework for product planning (analysis → PRD → architecture → epics).
- Greenfield: Fleet detects if BMAD commands are available and guides the engineer through BMAD's planning phases. If BMAD isn't installed, Fleet tells you.
- Brownfield: Fleet generates specs in BMAD-compatible format so the same build/review loop works regardless of where specs came from.
- Both use the same spec format, so fleet-build, fleet-review, and fleet-sync work identically on BMAD stories and Fleet-generated specs.
Skills
| Skill | Purpose |
|---|---|
fleet-init |
Universal entry point — detects repo state, routes to correct path |
fleet-discover |
Brownfield: map unknown repo (tech stack, structure, conventions) |
fleet-assess |
Brownfield: honest audit (classify complete vs stub vs missing) |
fleet-specgen |
Brownfield: generate BMAD-compatible specs from code gaps |
fleet-infra |
Set up test framework, CI pipeline, quality gate hooks |
fleet-sync |
Push repo state → Linear (issues) + Notion (docs) |
fleet-build |
TDD loop for one spec (write tests → implement → verify) |
fleet-review |
5-check verification (spec compliance, stubs, quality, security, regression) |
fleet-run |
Full orchestrator — parallel builds, verify, merge, sync, loop |
fleet-doctor |
Project health check — build, tests, spec accuracy, sync drift |
fleet-guard |
Process enforcement hooks (spec-before-code, auto-format) |
Agents
| Agent | Role |
|---|---|
fleet-builder |
Autonomous TDD implementation (worktree-isolated) |
fleet-reviewer |
Read-only 5-check verification |
fleet-security |
OWASP security audit |
fleet-sync-agent |
Scheduled Linear/Notion synchronization |
Principles
- Repo is truth. Linear and Notion are mirrors. One-way sync only.
- Specs before code. No implementation without acceptance criteria.
- TDD always. Tests written before implementation.
- Honest audits. Stubs are not implementations. Status must match reality.
- BMAD is upstream. Fleet never modifies BMAD framework files.