An engineering workflow for taking responsibility for an open GitHub pull request, which is a proposed set of code changes. It reviews the request, confirms findings, makes fixes on its branch, runs required web and desktop end-to-end tests, and reports the evidence without merging.

It helps review a pull request from its URL, check its files and automated checks, record confirmed prevention fixes, update the branch, and verify the result with end-to-end tests.

A pull request may contain unverified AI-generated mistakes or need work after its original author is unavailable. This provides a controlled way to inspect and correct it while keeping branches and workspaces separate.

pr-takeover — Skill for Claude Code by closedloop-ai/claude-plugins

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 closedloop-ai/claude-plugins --skill pr-takeover

Clone the repo

git clone --depth 1 https://github.com/closedloop-ai/claude-plugins

Made for: Claude Code.

Or install vibe, the plugin that ships this one along with the rest of its 17 skills, 17 agents.

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/closedloop-ai/claude-plugins/pr-takeover/github.svg)](https://agentmods.dev/skills/closedloop-ai/claude-plugins/pr-takeover)

Your own site

<a href="https://agentmods.dev/skills/closedloop-ai/claude-plugins/pr-takeover"><img src="https://agentmods.dev/badge/skills/closedloop-ai/claude-plugins/pr-takeover/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/closedloop-ai/claude-plugins/pr-takeover"><img src="https://agentmods.dev/badge/skills/closedloop-ai/claude-plugins/pr-takeover.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 64 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 2,317 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	$0.00064	$0.02317
Opus 5.5	$0.00026	$0.00927
Sonnet 5.5	$0.00013	$0.00463
Haiku 4.5	$0.00006	$0.00232

Measured 6d ago against content hash fe48ad8599e5, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

pr-takeover 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 6d 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.

plugins/vibe/skills/pr-takeover/SKILL.md · 51 lines

How it starts

The opening of the file, as written. The whole thing — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Engineer takeover of a vibe pull request

Input: one GitHub pull request URL. This skill works in a new chat with no vibe session, fleet board, local ticket, worktree, or decision table. Discover those facts from the URL and live systems. Do not require a fleet program. The engineer owns the final merge; this skill never merges or arms auto-merge.

This is an engineer PR takeover, not migration of a dead vibe session writer. A dead writer stays in the existing ../fleet-orchestrator/SKILL.md recovery path. This skill owns a new worktree and one writer for the PR source branch. Keep a separate claude-plugins worktree for the prevention worker. Neither worker edits the other's branch. Never reset, stash, or overwrite an existing worktree.

1. Resolve the PR before launching work

Parse the URL as a GitHub PR URL with owner, repository and numeric PR number. Refuse another URL shape before any write. Read the PR with gh pr view for state, base repository and branch, head repository and branch, head SHA, base SHA, author, body, review threads and checks. The live PR is the source of truth. Stop on a closed PR or missing push access and report the exact reason.
Read the linked ticket, PRD, human comments and applicable repository instruction files. A ticket body or a fleet instruction is not a human ruling. Preserve every settled behavior and flag-off path. Use closedloop-graph for ticket and code scope, then verify its material findings against the checkout.
Fetch the exact PR head and live base. Make a fresh worktree rooted at that head, with a local branch dedicated to this takeover and the PR's actual source branch and remote recorded. A source branch in a fork uses its head repository remote. Check the fetched head against the PR again before review and before every push. Keep one source writer in that worktree.
Read the repo's pnpm test:lanes selection for the PR diff. Read its E2E, browser and Desktop instructions, plus .github/pr-test-event-tiers.md before naming CI coverage. Do not infer a run from a workflow's existence.

Read the full file on GitHub · 51 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.

6d ago First seen · 51 lines · 64 tokens per session scan A fe48ad8599e5

Subscribe to this mod's changes

pr-takeover is a skill published in the GitHub repository closedloop-ai/claude-plugins (124 stars, last pushed today), licensed Apache-2.0. It adds 64 tokens to every session and 2,317 once invoked, about $0.0003 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 2026-11-03.

Claude Code Code review Git and workflow LLM agentic-ai ai-agents anthropic developer-tools

pr-takeover next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
pr-takeoverclosedloop-ai/claude-plugins	Skill	124	today	Apache-2.0	A	64	Claude Code
githubNousResearch/hermes-agent	Skill	262k	today	MIT	A	19	Claude Code, Codex
babysitmichael-denyer/pstack-claude	Skill	4.2k	today	MIT	A	61	Claude Code
make-pr-easy-to-reviewmichael-denyer/pstack-claude	Skill	4.2k	today	MIT	A	57	Claude Code
pr-creatormblode/agent-skills	Skill	191	today	MIT	A	55	Claude Code, Codex
review-this-branchno-human-ai/no_human	Skill	320	1mo ago	MIT	A	53	Claude Code
pr-describeMuhammadUsmanGM/claude-code-best-practices	Skill	109	16d ago	MIT	A	44	Claude Code