ClawSweeper is a maintenance bot for OpenClaw repositories that periodically reviews open issues and pull requests and identifies candidates for closure, repair, or guarded automerging. It is intended for repository maintainers running their own configured instance. The catalogue entries are skills and an instruction for its review and maintenance workflows.
Borrowing it
Nothing to install: this file belongs to openclaw/clawsweeper. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/openclaw/clawsweeper/main/AGENTS.mdgit clone --depth 1 https://github.com/openclaw/clawsweeperWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/openclaw/clawsweeper/agents-md)<a href="https://agentmods.dev/instructions/openclaw/clawsweeper/agents-md"><img src="https://agentmods.dev/badge/instructions/openclaw/clawsweeper/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/openclaw/clawsweeper/agents-md"><img src="https://agentmods.dev/badge/instructions/openclaw/clawsweeper/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.02581 | $0.02581 |
| Opus 5.5 | $0.01032 | $0.01032 |
| Sonnet 5.5 | $0.00516 | $0.00516 |
| Haiku 4.5 | $0.00258 | $0.00258 |
Grade A, and why
clawsweeper AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 195 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.MD
ClawSweeper is the conservative maintenance bot for OpenClaw repositories. Keep changes narrow, evidence-backed, and automation-safe.
AGENTS.md is the canonical repository-wide coding-agent policy. Task-specific
workflows under .agents/skills/ are opt-in procedures and do not override this
file. This repository does not currently maintain AGENT.md, CLAUDE.md, or
Cursor aliases; agents that do not discover AGENTS.md automatically must be
directed here rather than given a duplicated policy copy.
Structure
- Main code:
src/clawsweeper.ts. - Repair lane code:
src/repair/; canonical records live in the Cloudflare Worker, ledger/assets blobs live in R2, and remaining operational state lives inopenclaw/clawsweeper-state. - Tests:
test/*.test.tsandtest/repair/*.test.ts; add new coverage to the narrowest matching test file instead of growingtest/clawsweeper.test.ts. - Workflow:
.github/workflows/sweep.yml. - Vision and product boundaries:
VISION.md. - Explainer:
README.md; state/dashboard repo:../clawsweeper-state. - Changelog: one-bullet fragments in
changelog.d/, stitched intoCHANGELOG.mdat release time; seechangelog.d/README.md. - Open/reviewed records in the canonical Worker store:
records/<repo-slug>/items/<number>.md. - Archived records in the canonical Worker store:
records/<repo-slug>/closed/<number>.md. - Scratch/generated output:
.artifacts/,artifacts/,apply-report.json.
Preserve one flat items/ and closed/ report layout per repository slug. Do
not split reports into issue/PR subtrees.
Operating Model
- Review lane is proposal-only. It never closes GitHub items.
- Ordinary PRs receive one automatic completed review after ten quiet minutes.
Later pushes, edits, reopen events, and scheduled passes require a new explicit
review request. Failed or stale-source attempts can finish the first review.
clawsweeper:autofixandclawsweeper:automergeno longer change review admission; they only keep their close protection. - Maintainer-authored items and items labeled
maintainercreated from 2026-10-15 03:50 UTC are reviewed like everyone else's: a PR gets its one automatic review. Older maintainer items need an explicit@clawsweeper reviewrequest. Finish the work and validation first; request review only when it is needed. Agents must not request reviews automatically or repeat a queued/running request. An explicit review request stays open while the queue requeues its own review after source drift, in place or from the publisher still holding the request's publication; that retry continues the request. A finished or replaced request, later changes after the answer, and scheduled passes need a new request. - Apply lane mutates GitHub by syncing the durable Codex review comment and then closing only unchanged, high-confidence proposals.
- Repository-specific rules live in
src/repository-profiles.ts; ClawHub apply may close only PRs that are certainly implemented onmain. - Hosted review concurrency and batching belong to the shared exact-review queue. Broad sweeps only admit candidates; they do not start per-run shards or grant immediate-apply authority.
openclaw/clawsweeper-stateis the live status surface and operational state store; the Worker/R2 pair is authoritative for records, ledger, and assets. Check current Actions and the canonical owner before trusting local generated timestamps.- When asked about PRs outside
openclaw/clawsweeper, treat the task as monitoring/debugging how ClawSweeper workflows operate on that PR. Do not fix foreign PR branches directly; ClawSweeper repair/automerge workflows own those branch edits. - In
openclaw/openclaw,CHANGELOG.mdis release-owned. Do not ask contributor PR authors, repair workers, or automerge/autofix lanes to edit it during normal PR work. Preserve release-note context in PR bodies and commit messages instead. - In
openclaw/clawsweeperitself, record a user-visible change as one newchangelog.d/<pr-or-slug>.mdfragment holding a single bullet line. Do not editCHANGELOG.mdor other PRs' fragments in normal PR work; the release maintainer folds fragments in withpnpm run changelog:stitch, andpnpm run check:changelogvalidates them. - When referencing GitHub issues or PRs in user-facing output, always include
the full GitHub URL, not only
#12345.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago Changed · +4 lines · +60 tokens per session 4dd1f65c77ac
- 11d ago Changed · +3 lines · +47 tokens per session 4249b62e269b
- 13d ago Changed · +3 lines · +35 tokens per session d1eae78a3068
- 14d ago Changed · +8 lines · +126 tokens per session 27c353885004
- 15d ago Changed · +1 lines · +10 tokens per session a9f6d04aa82f
- 27d ago Changed · +9 lines · +169 tokens per session b9b0555ef16a
- 28d ago Changed · +1 lines · +18 tokens per session e32b0d9485a4
- 1mo ago Changed 324a525b1d5e
- 1mo ago Changed · +4 lines · +69 tokens per session a7f162d65762
- 1mo ago First seen · 162 lines · 2,047 tokens per session scan A 0ff005dafa53
clawsweeper AGENTS.md is an instructions file published in the GitHub repository openclaw/clawsweeper (2,004 stars, last pushed today), licensed MIT. It adds 2,581 tokens to every session, about $0.0103 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.