Skip to content

LFG Autonomous Pipeline

Takes a coding request from plan to pushed pull request with minimal supervision

Stars
25,850
Licence
MIT
Files
11 files

What it does

Acts as an orchestrator: when you explicitly ask for hands-off end-to-end work, it routes your request to other Compound Engineering skills (planning, implementation, debugging, review, simplification, shipping). For code changes it aims to finish with an open pull request, after implementing, simplifying, reviewing, and committing, and pushes without stopping for approval. It stops and reports only when a required step is blocked or an irreversible action falls outside what you granted, such as merging.

Great for

  • Developers who already trust the surrounding Compound Engineering skills and want an unattended run from a request to a pull request
  • Teams running scheduled or nested agent workflows that cannot pause for questions

Skip it if

  • Users who want to review each step: the SKILL.md itself points to ce-plan, ce-work, ce-debug or ce-commit-push-pr for that, because lfg pushes without waiting
  • Repositories where pushes or pull requests need human sign-off that the skill cannot see, since it can push and open PRs on its own

Try asking

  • “lfg add CSV export to the reports page”
  • “lfg fix the login redirect bug described in issue 412”
  • “lfg take plans/billing-retry.md end to end”
Step-by-step guide: how to use LFG Autonomous PipelineInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 11 original files, the MIT licence and an attribution note.

    lfg.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip lfg.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip lfg.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Be explicit that you want autonomous end-to-end work, since the skill only triggers on that or a direct lfg invocation; use ce-plan or ce-work for step-by-step review.
  • Make sure the project's branch, remote, and merge rules are set up before running, because pushes happen without a pause.
  • Read the generated pull request body for any recorded unapplied review findings or flagged decision conflicts before merging.

What's inside

  • references/debug-return.md
  • references/intake.md
  • references/next-work-handoff.md
  • references/plan-brief.md
  • references/review-followup.md
  • references/shipping.md
  • references/stage-routing.md
  • references/task-visibility.md
  • references/tracker-defer.md
  • references/work-return.md
  • SKILL.md

Redistributed unmodified from EveryInc/compound-engineering-plugin under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.