Official agent skill

Review Triage Phase

by prisma in prisma/orm

Runs the triage step of the review-framework loop: reads fetched PR review state, builds `review-actions.json`, validates it and renders `review-actions.md`.

OfficialApache-2.0Auto-check passedDevelopment

Install Review Triage Phase

skills CLI
$ npx skills add prisma/orm --skill review-triage-phase -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install prisma/orm review-triage-phase --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/prisma/orm.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills-contrib/review-triage-phase .claude/skills/review-triage-phase && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
review-triage-phase
GitHub stars
48k
Token cost
~995 tokens
SKILL.md length
385 words
Files
8 (incl. scripts)
Skills in repo
20
Repo updated
First seen
Licence
Apache-2.0

At a glance

Runs the triage step of the review-framework loop: reads fetched PR review state, builds `review-actions.json`, validates it and renders `review-actions.md`.

  • Works in 8 steps: Compute deterministic paths → Enforce artifact safety before… → Bootstrap canonical action scaffold from… → …
  • Turning fetched PR review comments into an ordered list of actions
  • SKILL.md covers Inputs, Preconditions, Behavior and Schema contract, plus 1 more section
  • Runs JavaScript scripts from its folder; calls node

What it does

You give it a PR URL and optionally an output directory, which otherwise defaults to a folder under `wip/reviews` named after the owner, repo and PR number. It expects `review-state.json` from the earlier fetch phase and points you to `/review-fetch-phase` when that file is missing. It first checks that review artifacts stay untracked, in either a git checkout or a Jujutsu workspace, then scaffolds `review-actions.json` with `scripts/bootstrap-review-actions.mjs` from the thread and review-body targets.

Triage itself is delegated to `agents/review-triager.md`, whose output must be valid v2 JSON, use node IDs only, keep actions in order, give every deferred action a Linear issue identifier and keep replies on review threads actionable as thread replies. Bundled reviews, such as a bot's combined nitpicks and outside-diff comments, must be split into separate items and never dismissed wholesale. Further scripts validate and render the actions file as markdown.

When your agent uses it

  • Turning fetched PR review comments into an ordered list of actions
  • Splitting a bundled bot review into individual action items
  • Planning which review comments to fix now and which to defer to Linear

Example prompts

  • “Triage the review comments on this PR and write review-actions.md without changing any code.”
  • “The fetch phase finished for my PR. Build review-actions.json and validate it.”
  • “Split the bundled bot review on this PR into separate actions and mark the ones to defer.”

Requirements

  • Node to run the bundled `.mjs` scripts
  • Output of the review fetch phase (`review-state.json`)
  • A Linear workspace for tracking deferred actions

Workflow steps

8 steps, taken from the first numbered list in SKILL.md.

  1. Compute deterministic paths
  2. Enforce artifact safety before generation (the artifacts must stay untracked). In a git checkout the guard asks git directly; in a Jujutsu…
  3. Bootstrap canonical action scaffold from review state
  4. Delegate triage to
  5. Require triager output contract
  6. Create Linear tracking for deferred work
  7. Validate canonical actions JSON and enforce completed triage decisions
  8. Render markdown from canonical actions JSON

What it can do on your machine

Read from SKILL.md and the folder at commit 19bbe0e. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 5 files in scripts/ (JavaScript), which the agent can run.

    Shell commands in SKILL.md call:

    • node

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Review Triage Phase loads about 995 tokens when it runs. Until then it costs about 49 tokens; SKILL.md has 385 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~49
When it runs · the whole SKILL.md, loaded when a task matches
~995

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from prisma/orm at commit 19bbe0e, republished under its Apache-2.0 licence (© prisma). 385 words, ~995 tokens.

Download SKILL.mdSave it as .claude/skills/review-triage-phase/SKILL.md (or your agent's skills folder). This skill also uses 7 other files; get the full folder from GitHub.
name
review-triage-phase
description
Produces canonical review actions from fetched review state and renders action markdown. Use when the user wants only triage/action-planning for the review-framework workflow.
argument-hint
[pr-url] [output-dir]

Review Triage Phase

Run only the triage phase of the review-framework loop:

read current review state artifacts, bootstrap canonical review-actions.json, update triage fields, validate, then render review-actions.md.

Run commands from this skill directory. All script paths below are relative to it.

Inputs

  • Required:
    • PR URL
  • Optional:
    • output directory

If output directory is omitted, derive:

wip/reviews/<owner>_<repo>_pr-<number>/

Preconditions

Expected inputs in output dir:

  • <output-dir>/review-state.json
  • <output-dir>/review-targets.json (optional)
  • <output-dir>/review-state.md (optional)
  • <output-dir>/summary.txt (optional)

If review-state.json is missing, instruct user to run:

  • /review-fetch-phase <PR_URL> [output-dir]

Note:

  • review-actions.json scaffolding is generated from review-state.targets, taking the thread and review-body targets. Issue-comment targets are captured in review-state.json but get no scaffolded action.
  • Compound review bodies: A single pull_request_review target may contain multiple distinct findings (e.g., CodeRabbit bundles "outside diff range" comments, actionable comments, and nitpicks into one review body). The triager must decompose these into individual action items during triage — never blanket-dismiss a review body without reading its content.

Behavior

  1. Compute deterministic paths:
    • <output-dir>/review-state.json
    • <output-dir>/review-actions.json
    • <output-dir>/review-actions.md
  2. Enforce artifact safety before generation (the artifacts must stay untracked). In a git checkout the guard asks git directly; in a Jujutsu workspace with no git directory it accepts a directory under the workspace root's ignored wip/ tree:
bash
node ../review-fetch-phase/scripts/guard-review-artifacts-ignored.mjs --dir <output-dir>
  1. Bootstrap canonical action scaffold from review state:
bash
node ./scripts/bootstrap-review-actions.mjs --in <output-dir>/review-state.json --out <output-dir>/review-actions.json
  1. Delegate triage to:
  • ./agents/review-triager.md
  1. Require triager output contract:
    • review-actions.json is valid v2
    • targets use node IDs only
    • actions remain intentionally ordered
    • every defer action includes a Linear issue identifier
    • will-address actions for review_thread targets must remain actionable via thread replies (not PR reviews)
  2. Create Linear tracking for deferred work:
    • create one or more Linear tickets (group related deferred comments)
    • attach each deferred action to a Linear issue ID (linearIssue)
    • ensure the deferred thread reply includes that Linear ticket URL
  3. Validate canonical actions JSON and enforce completed triage decisions:
Show full SKILL.md (88 more words)Show less
bash
node ./scripts/validate-review-actions.mjs --in <output-dir>/review-actions.json --require-final
  1. Render markdown from canonical actions JSON:
bash
node ./scripts/render-review-actions.mjs --in <output-dir>/review-actions.json --out <output-dir>/review-actions.md

Schema contract

  • review-actions.json is canonical and must be schema version 2.
  • No backward compatibility is provided for v1 artifacts.
  • The triager edits scaffolded actions in place instead of reconstructing the entire file.
  • Triage phase is only complete when there are zero triage_pending decisions.
  • Review artifacts are generated files and must remain untracked in git.
  • defer decisions require Linear tracking (linearIssue) and a thread reply with that link.

Output to user

Return artifact paths:

  • review-actions.json
  • review-actions.md

Suggest next step:

  • /review-implement-phase <PR_URL> [output-dir]

© prisma, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 7 other files (scripts) in skills-contrib/review-triage-phase of prisma/orm.

  • SKILL.md
  • agents/review-triager.md
  • package.json
  • scripts/bootstrap-review-actions.mjs
  • scripts/render-review-actions.mjs
  • scripts/render-review-actions.test.mjs
  • scripts/review-artifacts.mjs
  • scripts/validate-review-actions.mjs

Open the folder on GitHubat commit 19bbe0e

Compare with similar skills

Review Triage Phase next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Review Triage Phase compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Review Triage Phase this skillprisma/orm48k—~995Automated safety check: PassApache-2.0
PR Triagertk-ai/rtk83k—~2.5kAutomated safety check: NotesApache-2.0
Adopt PR Branch Contextpydantic/pydantic-ai-harness952—~1.8kAutomated safety check: PassMIT
Open Code Review CLIalibaba/open-code-review45k—~3.1kAutomated safety check: PassApache-2.0
Understand Diff AnalysisEgonex-AI/Understand-Anything86k—~1.4kAutomated safety check: PassMIT
Code Reviewflutter/flutter179k—~1.4kAutomated safety check: PassBSD-3-Clause

Similar skills

  • PR Triage

    rtk-ai/rtk

    Audits a repository's open pull requests, deep-reviews chosen ones and drafts review comments that are only posted after you approve them.

    83k GitHub stars~2.5k tokensUpdated yesterday
    DevelopmentAuto-check: notes
  • Adopt PR Branch Context

    pydantic/pydantic-ai-harness

    Official

    Fills in issue-brief.md and pr-decisions.md for an existing pull request, so you can pick up a PR mid-flight with its linked issue and past review decisions summarized.

    952 GitHub stars~1.8k tokensUpdated 7 days ago
    DevelopmentAuto-check passed
  • Open Code Review CLI

    alibaba/open-code-review

    Runs the ocr command-line tool to review Git changes, a commit or a branch comparison with an AI model, returning line-level comments and optionally applying fixes.

    45k GitHub stars~3.1k tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Understand Diff Analysis

    Egonex-AI/Understand-Anything

    Reads your git changes or a pull request against a prebuilt knowledge graph of the project to explain what changed, which components are affected and what is risky.

    86k GitHub stars~1.4k tokensUpdated today
    DevelopmentAuto-check passed
  • Code Review

    flutter/flutter

    Performs a comprehensive, multi-step code review of pull requests or local code changes, using iterative refinement (generation, critique, synthesis) to ensure high-quality, actionable feedback.

    179k GitHub stars~1.4k tokensUpdated today
    DevelopmentAuto-check passed
  • Knowledge Graph PR Review

    tirth8205/code-review-graph

    Reviews a pull request or branch diff with a code knowledge graph and produces a structured review that includes blast-radius analysis.

    32k GitHub stars~452 tokensUpdated 3 days ago
    DevelopmentAuto-check passed

More from prisma/orm

All 20 skills in this repo
  • Official

    Runs a loop on a GitHub pull request: fetch review state, triage comments into actions, implement them and resolve threads, repeating until nothing actionable is left.

    48k GitHub stars~2.2k tokensUpdated today
    Auto-check passed
  • Official

    Fetches a pull request's canonical review state as JSON, validates it, and renders markdown, a text summary and triage target files from it using bundled scripts.

    48k GitHub stars~767 tokensUpdated today
    Auto-check passed
  • Official

    Implements triaged pull request review actions, commits focused fixes, posts status replies on GitHub and resolves the threads.

    48k GitHub stars~1.7k tokensUpdated today
    Auto-check passed
  • Official

    Replaces a plain TypeScript union plus switch statements with frozen subclasses and a visitor interface when several places dispatch on the same variants.

    48k GitHub stars~830 tokensUpdated today
    Auto-check passed
  • Official

    Guides an outside contributor through opening a prisma/orm pull request from a fork that follows CONTRIBUTING.md and passes review on the first round.

    48k GitHub stars~2.3k tokensUpdated today
    Auto-check passed
  • Official

    Creates a GitHub pull request for the Prisma repo with a Linear-ticket-prefixed title and a decision-led narrative description, inferring the ticket from context.

    48k GitHub stars~4.6k tokensUpdated today
    Auto-check passed

Works with

Questions about Review Triage Phase

What does Review Triage Phase do?

Runs the triage step of the review-framework loop: reads fetched PR review state, builds `review-actions.json`, validates it and renders `review-actions.md`. You give it a PR URL and optionally an output directory, which otherwise defaults to a folder under `wip/reviews` named after the owner, repo and PR number.json` from the earlier fetch phase and points you to `/review-fetch-phase` when that file is missing.

When should I use Review Triage Phase?

Review Triage Phase fits situations like: turning fetched PR review comments into an ordered list of actions; splitting a bundled bot review into individual action items; planning which review comments to fix now and which to defer to Linear.

How do I install Review Triage Phase in Claude Code?

Run `npx skills add prisma/orm --skill review-triage-phase -a claude-code`. Or copy the skill folder (skills-contrib/review-triage-phase in prisma/orm) into .claude/skills/review-triage-phase in your project. Claude Code loads it when a task matches its description.

How do I install Review Triage Phase in Codex?

Run `npx skills add prisma/orm --skill review-triage-phase -a codex`. Or copy the skill folder (skills-contrib/review-triage-phase in prisma/orm) into .agents/skills/review-triage-phase in your project. Codex loads it when a task matches its description.

Can I use Review Triage Phase in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add prisma/orm --skill review-triage-phase -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/review-triage-phase, .gemini/skills/review-triage-phase, .github/skills/review-triage-phase and .opencode/skills/review-triage-phase in your project.

What does Review Triage Phase need to run?

Going by SKILL.md and its folder, Review Triage Phase needs JavaScript for the scripts in its folder and the command-line tools its instructions call (node). Our summary lists: Node to run the bundled `.mjs` scripts; Output of the review fetch phase (`review-state.json`); A Linear workspace for tracking deferred actions.

Does Review Triage Phase access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Review Triage Phase safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Review Triage Phase use?

Review Triage Phase is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Review Triage Phase use?

About 995 tokens (SKILL.md is roughly 4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Review Triage Phase?

Skills that share tags, products or a category with Review Triage Phase: PR Triage (rtk-ai/rtk, 83k stars), Adopt PR Branch Context (pydantic/pydantic-ai-harness, 952 stars), Open Code Review CLI (alibaba/open-code-review, 45k stars) and Understand Diff Analysis (Egonex-AI/Understand-Anything, 86k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Review Triage Phase?

prisma (a GitHub organization, an official publisher) maintains it in prisma/orm, which has 47,696 GitHub stars. The repository holds 20 skills in this directory. The repository was last updated on October 9, 2026.

Source: prisma/orm on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.