How do I install the Pr Review skill?

Run npx skillmds@latest add beskars-code/pr-review in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Pr Review safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Pr Review?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Pr Review free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Pr Review?

beskars-code (@beskars-code) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Pr Review by beskars-code — Agent Skill

Name: Pr Review
Author: beskars-code

Pr Review

Stateless review of an open GitHub PR. Standards hard-require /taste KISS + named principles (including honest names, trust the server, types tell the truth, and naming alignment) and /architecture, plus a correctness hunt and PR extras (body vs diff, migrations, breaking API). Triage historical finding threads, draft one-topic comments, post only after one publish decision.

beskars-code 0c7ad68 3 files · 13.0 KB Updated Sep 17, 2026

File contents

PR Review

Must read: ../pack-shared/standards.md. Read /taste and /architecture doctrines this turn before adjudicating Standards. Do not skip.

Ask style: ../pack-shared/asking.md

This skill is a user start. Do not nest it under /task.

Read doctrine.md and reference.md. The shared review contract owns review depth, evidence, the review output fence, finding records, and severity mapping. The shared execution context owns stateless authority and handoff rules.

Standards: always Read /taste and /architecture this turn. Apply code-review:blocker-vs-follow-up, code-review:naming-alignment, the Correctness hunt, and pr-review:* extras on initial and full-rescan (and on newly introduced follow-up surface). Do not run a Design axis.

Process

Pin the PR fixed point (headSha) and load its body, commits, diff, all inline review comments, and all review threads, including resolved threads and every prior review page. Record previousReviewedHead from the last review this skill completed on this PR when available (from chat or the latest review commit association).
Dispatch pr-reviewer Tasks per ../pack-shared/subagents.md. Feed the what (open GitHub PR, injected plan, bars). The worker owns how to hunt. The parent reviews Completions and owns the publish question. Do not solo non-trivial PR review labor.
With no prior finding thread, run the shared contract's initial review.
On every follow-up, complete Pass A first (doctrine Output), then Pass B.
Use full-rescan only when the user explicitly requests it or materially expands the review scope.
Show drafts, ask one publish question, apply the stale-head guard, then post only after approval.

---name: pr-reviewdescription: Stateless review of an open GitHub PR. Standards hard-require /taste KISS + named principles (including honest names, trust the server, types tell the truth, and naming alignment) and /architecture, plus a correctness hunt and PR extras (body vs diff, migrations, breaking API). Triage historical finding threads, draft one-topic comments, post only after one publish decision.---# PR Review**Must read:** [../pack-shared/standards.md](../pack-shared/standards.md). Read `/taste` and `/architecture` doctrines this turn before adjudicating Standards. Do not skip.**Ask style:** [../pack-shared/asking.md](../pack-shared/asking.md)This skill is a user start. Do not nest it under `/task`.Read [doctrine.md](doctrine.md) and [reference.md](reference.md). The shared[review contract](../pack-shared/review-contract.md) owns review depth,evidence, the review output fence, finding records, and severity mapping. Theshared [execution context](../pack-shared/execution-context.md) owns statelessauthority and handoff rules.**Standards:** always Read `/taste` and `/architecture` this turn. Apply`code-review:blocker-vs-follow-up`, `code-review:naming-alignment`, theCorrectness hunt, and `pr-review:*` extras on initial and full-rescan (and onnewly introduced follow-up surface). Do not run a Design axis.## Process1. Pin the PR fixed point (`headSha`) and load its body, commits, diff, all inline review comments, and all review threads, including resolved threads and every prior review page. Record `previousReviewedHead` from the last review this skill completed on this PR when available (from chat or the latest review commit association).2. Dispatch `pr-reviewer` Tasks per [../pack-shared/subagents.md](../pack-shared/subagents.md). Feed the what (open GitHub PR, injected plan, bars). The worker owns how to hunt. The parent reviews Completions and owns the publish question. Do not solo non-trivial PR review labor.3. With no prior finding thread, run the shared contract's `initial` review.4. On every follow-up, complete **Pass A** first (doctrine Output), then **Pass B**.5. Use `full-rescan` only when the user explicitly requests it or materially expands the review scope.6. Show drafts, ask one publish question, apply the stale-head guard, then post only after approval.

Frequently asked questions