loop-engineering README

The project's README from github.com/cobusgreyling/loop-engineering. As one markdown file: README.md.

Loop Engineering

Showcase GitHub stars loop npm last triage MIT

Loop Engineering logo

Stop prompting. Design the loop. Get a score.

Loop Engineering — design the system that prompts your agents

Start in 5 minutes · Quickstart · What do you want to do? · Refactor path · 中文 · Pattern picker 

npx @cobusgreyling/loop init . --pattern daily-triage --tool claude
npx @cobusgreyling/loop doctor .

--tool defaults to claude if you omit it. Swap for grok, codex, or opencode. Week one is report-only.

Loop Ready score climbing

You design a system that discovers work, hands it to agents, verifies results, and persists state — instead of typing the next prompt yourself.

New here? Quickstart · Refactor a project · Thin GitHub Action loop (no STATE.md required)

What do you want to do?

This is a pattern library for operating agents around a codebase. It is not a “rewrite the module” button.

I want to… Start here
Keep the repo healthy (issues, CI, deps) Quickstart → daily-triage
Land and babysit PRs PR Babysitter
Run a loop in GitHub Actions with almost no files Thin loop
Ship a feature or refactor (todos → small PRs) Refactor / change path

Full jobs table: docs/jobs.md.

Patterns

Pattern Cadence Week 1 Cost
Daily Triage 1d–2h L1 report Low
Thin loop event + 1d L1 snapshot Very low
PR Babysitter 5–15m L1 watch High
CI Sweeper 5–15m L2 cautious Very high
Dependency Sweeper 6h–1d L2 patch-only Medium
Changelog Drafter 1d or tag L1 draft Low
Post-Merge Cleanup 1d–6h L1 off-peak Low
Issue Triage 2h–1d L1 propose-only Low

Interactive picker: showcase. Index: patterns/registry.yaml.

Getting started

npx @cobusgreyling/loop init . --pattern daily-triage --tool claude
npx @cobusgreyling/loop doctor .
npx @cobusgreyling/loop cost --pattern daily-triage --level L1

Empty repo → first STATE.md walkthrough: scripts/empty-to-state-demo.sh.

Roll out L1 report → L2 assisted → L3 unattended only after the verifier has been right for a week. See loop-design-checklist.

Loop Ready now weights recent runs harder than files on disk. A 30-day-old STATE.md is not L3.

Examples by tool

Claude Code (including the plugin) · Grok · Codex · OpenClaw · Opencode · GitHub Actions

Operating & safety

Failure modes · Anti-patterns · Safety · Operating loops · Stories (wins and failures)

Loop engineering amplifies judgment. Token costs can explode. Unattended loops make unattended mistakes. Read what the loop ships.

Help wanted

Docs and small PRs: first response within 48 hours.

See CONTRIBUTING.md.

CLI packages and companion repos (optional — not required for week one)

Unified front door: npx @cobusgreyling/loop (init · doctor · status · audit · cost). Older packages (loop-init, loop-audit, …) stay supported.

Companions exist for later: memory-engineering, harness-foundry, outerloop, fleet-engineering, goal-engineering. Do not add them until a loop has actually run.

Sources

License

MIT

loop-engineering (cobusgreyling/loop-engineering)

Searching Database…

11.6K ·GitHub
Loop-engineering
What is this?Loop Engineering is a pattern library and set of CLI tools for designing automated loops where AI coding agents discover work, act on it, verify the result, and persist state in the repo, instead of you typing the next prompt yourself.
Is this a framework or just a bunch of CLI tools?It is mainly a documented pattern library (Daily Triage, PR Babysitter, CI Sweeper, etc.) plus companion CLI tools like loop-audit, loop-init and loop-cost, and GitHub Actions templates that implement those patterns — not a one-click code rewriter.
Which coding agents does it work with?Examples and a --tool flag cover Claude Code (including a plugin), Grok, Codex, OpenClaw, and Opencode; Claude is the default if you don't specify one.
Does it run my agents fully unattended out of the box?No — you're meant to roll out gradually from L1 (report-only) to L2 (assisted) to L3 (unattended) only after the verifier has been proven right for a week; an old STATE.md file doesn't count as being ready for L3.
What patterns does the library include?Eight documented jobs cover common repo maintenance: Daily Triage, Thin loop, PR Babysitter, CI Sweeper, Dependency Sweeper, Changelog Drafter, Post-Merge Cleanup and Issue Triage, each with a suggested cadence, starting safety level, and rough cost.
What CLI tools come with the project?Older standalone tools like loop-init, loop-audit, loop-cost, loop-sync, loop-context and loop-gate are still supported, and are now unified under a single npx @cobusgreyling/loop command.
Why was loop engineering created?It formalizes the idea of designing the system around an AI coding agent instead of writing one-off prompts, building on ideas from Addy Osmani and Boris Cherny, and is cited as the community reference implementation in a 2026 academic paper on agent loop adoption.
What's the biggest risk when running these loops?The README warns that loop engineering amplifies judgment, so token costs can explode and an unattended loop can make unattended mistakes — you're expected to actually read what the loop ships.
What are similar projects?agents, pstack-claude, open-swe, pstack, qwen-code, jevmem, claude-code, headroom
Which topics is it in?loop-engineering, claude-code-plugin, devtools, grok, ai-coding, coding-agents, github-actions, github, workflow-automation, agent-orchestration, prompt-engineering, agentic-ai
Creator
cobusgreyling
Size
6.3 MB
Stars
11.6K
Updated
CommitOct 18Releasev1.6.0Jul 20
Age
4mo
Commits · 90 Days