A workflow for reviewing every eligible open GitHub pull request, a proposed code change, and taking confirmed fixes through passing CI, the automated build and test checks.

Use it to list open pull requests, review them, verify reported problems, apply appropriate fixes, run tests, and merge branches when instructed.

It organizes review, evidence checking, editing, testing, and merging so findings are not applied blindly. It can also stop before merging.

pr-sweep — Skill for Claude Code by Chemaclass/agnostic-ai

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent

npx skills add Chemaclass/agnostic-ai --skill pr-sweep

Clone the repo

git clone --depth 1 https://github.com/Chemaclass/agnostic-ai

Made for: Claude Code.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/chemaclass/agnostic-ai/pr-sweep/github.svg)](https://agentmods.dev/skills/chemaclass/agnostic-ai/pr-sweep)

Your own site

<a href="https://agentmods.dev/skills/chemaclass/agnostic-ai/pr-sweep"><img src="https://agentmods.dev/badge/skills/chemaclass/agnostic-ai/pr-sweep/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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/chemaclass/agnostic-ai/pr-sweep"><img src="https://agentmods.dev/badge/skills/chemaclass/agnostic-ai/pr-sweep.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 47 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,098 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

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	$0.00047	$0.01098
Opus 5.5	$0.00019	$0.00439
Sonnet 5.5	$0.00009	$0.00220
Haiku 4.5	$0.00005	$0.00110

Measured 4d ago against content hash f2631be65508, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

pr-sweep 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 4d 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.

.agnostic-ai/skills/pr-sweep/SKILL.md · 80 lines

How it starts

The opening of the file, as written. The whole thing — 80 lines — stays where its author put it; the contents beside it link to each section on GitHub.

PR sweep

Take every open PR from review to merge without asking between steps. With PR numbers, sweep only those. With --no-merge, stop at green PRs.

1. List

gh pr list --state open --json number,title,headRefName,additions,deletions,author

Skip PRs by other authors and drafts; report them. Empty list: stop.

2. Review

Run two reviews on the PR branch, in parallel:

The code-reviewer agent for Go diffs (other diffs against the project rules). Keep its Opus model; never pass a cheaper override.
An adversarial review that challenges the design, not only the lines: /codex:adversarial-review --base origin/main where the Codex plugin is installed (Claude Code). Without it, run a second reviewer pass told to question the approach, its assumptions, and how it fails in real use.

Read every result before editing. After each round of fixes, run the adversarial review again until it approves or every remaining finding is rejected in step 4.

3. Verify each finding

A finding is a claim, not a task. Before acting on one:

Reopen the file and line it cites on the PR head.
Reproduce behavior claims with a built binary or a failing test.
Check vendor claims against the vendor page, not memory.

Keep what holds. Record what does not, with the reason.

4. Apply

Check out the PR branch in the main checkout (worktree agents cannot run git while the RTK hook rewrites it) and pull. For each confirmed finding:

Fix it on the branch. Test first for behavior changes: the new test must fail on the old code.
A finding that needs a different design or touches other targets gets its own issue (gh issue create --assignee Chemaclass with a label). Do not widen the PR.
A finding you reject after verification is answered, not ignored.

Then post one PR comment listing the commit that applied the findings, the issues filed, and each rejection with its reason.

5. Gate

Before every push, all green. Check each command by its own exit code; a pipe such as make ci-local | grep FAIL | head exits 0 when the gate fails.

Read the full file on GitHub · 80 lines

Changes

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.

4d ago Changed f2631be65508
6d ago Changed 7f5f7c22b8c1
17d ago Changed · +7 lines · +24 tokens per session 996b3d3db3ef
20d ago Changed · -3 lines 57222c6a65df
22d ago First seen · 76 lines · 23 tokens per session scan A fe649e092bfd

Subscribe to this mod's changes

pr-sweep is a skill published in the GitHub repository Chemaclass/agnostic-ai (25 stars, last pushed today), licensed MIT. It adds 47 tokens to every session and 1,098 once invoked, about $0.0002 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 2026-09-24.

Code review Git and workflow agentic-ai agents-md agnostic-ai claude codex gemini

eg-precommit-review

Run a Codex-native pre-commit independent-review loop on pending changes.

vshvedov/elephant-goldfish · 20 tokens

github-pr-workflow

Prepare a GitHub pull request from a feature branch — branch hygiene, commit shape, title/body, verification notes, screenshots for UI work, and replies to review comments.

paperclipai/paperclip · 39 tokens

agent-swarm-pr

Agent skill for swarm-pr - invoke with $agent-swarm-pr.

ruvnet/ruflo · 18 tokens

agent-pr-manager

Agent skill for pr-manager - invoke with $agent-pr-manager.

ruvnet/ruflo · 16 tokens

github-automation

GitHub workflow automation, PR management, issue tracking, and code review coordination. Integrates with GitHub Actions and repository management. Use when: PR creation, code review, issue management, release automation, workflow setup. Skip when: local-only changes, non-GitHub repositories.

ruvnet/ruflo · 61 tokens

babysit

Watch an open PR — fix failing CI, handle the straightforward review comments, and drive it to a mergeable state. Claude Code analog of Cursor's built-in /babysit. Use after opening a PR when the user wants the agent to shepherd it without re-prompting.

michael-denyer/pstack-claude · 61 tokens