{"slug": "pyor-review-claude-skill-for-human-code-review", "title": "/pyor:review – Claude skill for human code review", "summary": "Pyor released /pyor:review, a Claude Code plugin and cross-agent skill that opens working changes as a local pre-PR review in the Pyor desktop app before a PR exists. The tool runs a panel of fresh sub-agents with no context from the calling session across three lenses — correctness, simplicity, and security — to produce file groups and inline hints, and requires no API key or in-app model run. It works with Claude Code, Codex, Cursor, and other coding agents, and supports optional grouping intents including walkthrough, custom, and plain.", "body_md": "Agent tooling for [Pyor](https://pyor.review), the native home for GitHub code\nreview. Works with **Claude Code, Codex, Cursor, and any coding agent.**\n\n[Watch the 5-minute demo](https://youtu.be/bTvkZOHDKP4): /pyor:review in Claude Code, notes in Pyor, fixes back in the branch.\n\nOpen your working changes as a **local pre-PR review** in the Pyor desktop app,\nbefore a PR exists. `/pyor:review` has a panel of fresh sub-agents read the diff\nand hand Pyor two things: **file groups** (the changes sorted into labelled\nfolders, important first) and **inline hints** (short pointers to risks worth a\nlook). No API key, no in-app model run.\n\nThe AI here is deliberately quiet: it **points you at what matters** and gets out\nof the way. No summaries, no walls of generated text to wade through, just\ngroups to orient you and the occasional hint on a line worth a second look. You\nread the actual code.\n\n- \n**`/pyor:review`** , the primary command.**Authors an AI review by default.** A panel of**fresh sub-agents** produces the file grouping + inline hints from\nthe diff, then hands them to Pyor to render. The reviewers run with**no\ncontext from the calling session** (they see only the code, so the review\nisn't biased by whoever wrote it) across three lenses: correctness,\nsimplicity, and security. After you read and comment in Pyor, click**\"Send to\nClaude\"** to push your comments straight back into the session that opened it.Takes an optional grouping intent: \n\n```\n/pyor:review                      # importance: signal vs noise, important first (default)\n/pyor:review walkthrough          # order the groups as a reading path through the change\n/pyor:review custom \"by feature\"  # your own grouping instruction\n/pyor:review plain                # skip the AI panel, just open the diff (alias: fast)\n```\n\nBoth open your working changes as a local pre-PR review, ready to read before a PR exists; re-running reopens the same review (idempotent).\n\n**Claude Code**, install the plugin:\n\n```\n/plugin marketplace add Pyor-review/pyor-cli\n/plugin install pyor@pyor\n```\n\n**Any other agent** (Codex, Cursor, Amp, Gemini CLI, Windsurf, Zed, …), install\nthe skill with [`skills`](https://github.com/vercel-labs/skills), which mirrors\nit into your agent's format:\n\n```\nnpx skills add https://github.com/Pyor-review/pyor-cli/tree/main/plugins/pyor/skills/pyor-review\n```\n\nEither way, then run `/pyor:review` (or invoke the `pyor-review` skill) from any\nrepository. Both drive the same [`pyor-review`](#the-pyor-review-cli) flow.\n\n- The **Pyor desktop app** . It registers the`pyor://` URL scheme, renders the\nreview, and exports the review context the AI panel writes against. Grab it at[pyor.review](https://pyor.review) ;`/pyor:review` hands you the install\ncommand if it's missing.\n- `git` and`node` on your PATH.\n- `gh` (the GitHub CLI) is optional, used only when a command needs GitHub data.\n\nBoth the Claude plugin and the cross-agent skill drive one small, agent-neutral\nCLI that does the deterministic parts: git resolution, the working-tree\nrevision token (which must match what the app computes so aids cache-hit), the\n`pyor://` deep link, and the `~/.pyor` feedback channel. The agent only supplies\nthe analysis (grouping + hints).\n\nRun it without installing:\n\n```\nnpx -y pyor-review prepare\nnpx -y pyor-review --selftest\n# or install it: npm install -g pyor-review\n# no-install fallback before it's on npm:\n#   npx -y --package github:Pyor-review/pyor-cli pyor-review prepare\n```\n\nSubcommands: `prepare` (resolve repo/head/base + a deterministic session id +\nrevision + the exported review context), `open` (fire the deep link with the\naids), `wait` (park for \"Send to Claude\" feedback). `pyor-local-review` is the\nplain, no-AI open.\n\nThe scripts are self-contained and repo-agnostic, sharing\n`plugins/pyor/scripts/lib.mjs`:\n\n```\nnode plugins/pyor/scripts/pyor-local-review.mjs --print     # print the deep link, do not open\nnode plugins/pyor/scripts/pyor-ai-review.mjs prepare        # resolve repo/head/base + revision + context\nnode plugins/pyor/scripts/pyor-ai-review.mjs --selftest     # run the built-in assertions (incl. session-id determinism)\n```\n\n", "url": "https://wpnews.pro/news/pyor-review-claude-skill-for-human-code-review", "canonical_source": "https://github.com/pyor-review/pyor-cli", "published_at": "2026-09-16 13:48:40+00:00", "updated_at": "2026-09-16 14:14:28.642608+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "developer-tools", "ai-products"], "entities": ["Pyor", "Claude Code", "Codex", "Cursor", "Amp", "Gemini CLI", "Windsurf", "Zed"], "alternates": {"html": "https://wpnews.pro/news/pyor-review-claude-skill-for-human-code-review", "markdown": "https://wpnews.pro/news/pyor-review-claude-skill-for-human-code-review.md", "text": "https://wpnews.pro/news/pyor-review-claude-skill-for-human-code-review.txt", "jsonld": "https://wpnews.pro/news/pyor-review-claude-skill-for-human-code-review.jsonld"}}