teamcraft-jcg:pr-review
Conduct a pull request review without leaving Claude Code. Access the full PR diff, implementation context, and discussion thread. Leave review comments, respond to threads, approve, and merge — all without opening the GitHub UI.
What this skill does
## Goal A reviewer conducts a complete pull request review — reading the diff, understanding the implementation in context, engaging with existing discussion threads, leaving review comments, and deciding to approve or request changes — entirely within Claude Code, without opening the GitHub UI. ## Hard Constraints - The PR number comes from `$ARGUMENTS`. If not provided, ask. - Never auto-approve or auto-merge. Every consequential action — approving, merging, requesting changes — requires explicit instruction from the reviewer. - Use git (Bash) to understand codebase context beyond what the PR diff alone shows. The diff tells what changed; the codebase tells why it matters. - Human is the authority on what merges. Surface findings and engage with the reviewer. Do not decide. - **All GitHub interactions go through the gh CLI only.** All PR operations run via Bash: `gh pr view [PR-number]`, `gh pr diff [PR-number]`, `gh pr view [PR-number] --comments`, `gh pr comment [PR-number] --body "..."`, `gh pr review [PR-number] --approve`, `gh pr merge [PR-number]`, `gh pr edit [PR-number]`. - **Thread resolution is not available via gh CLI.** GitHub has no CLI equivalent for resolving review threads. After the review is complete, instruct the reviewer to resolve addressed threads in the GitHub UI manually. This is a known capability gap, not an error. ## Identify the Project Read `.teamcraft/project.md` if it exists — it contains the GitHub repo (owner/repo). If not, run `git remote -v` via Bash to identify the GitHub repo from the remote URL. No project listing tool is needed. ## Load the PR Fetch the PR details via `gh pr view [PR-number] --json number,title,body,headRefName,baseRefName,state,isDraft` via Bash. Read it completely: title, description, source branch, target branch, state, linked Jira issue key (from description), and any acceptance criteria the description contains. If the PR description references a Jira issue key (e.g., PROJ-42), note it for context — the original requirements and acceptance criteria belong in the reviewer's picture. Fetch the diff via `gh pr diff [PR-number]` via Bash. Fetch existing comments via `gh pr view [PR-number] --comments` via Bash. Read both in full before presenting anything to the reviewer. ## Orient the Reviewer Locally Check the current local branch. If the reviewer is not already on the PR's source branch, offer to check it out: > "You're currently on `[branch]`. Want me to check out `[source-branch]` so you can view the code locally in your editor?" Wait for their answer. Check it out if they say yes. A reviewer seeing the code in their editor alongside the diff has a better picture than the diff alone. ## Understand the Implementation in Context Read the changed files in the local codebase alongside the diff. The diff shows what changed; the surrounding code shows what it changed relative to. Use `git log` on the source branch via Bash to understand the commit sequence and how the implementation evolved. Look for what the automated analysis (code health and security review, if it ran) already flagged. CLAUDE.md and `.claude/rules/` contain team conventions — read them to understand what this codebase expects and whether the implementation follows it. ## Present the PR to the Reviewer Synthesize the PR for the reviewer — not a raw diff dump. Cover: **What this PR delivers** — a clear statement of what was built, referencing the linked Jira issue and acceptance criteria. **What changed** — the meaningful changes, grouped by area or concern. Not a line-by-line recitation. The reviewer can drill into specifics; start with the picture. **Existing discussion** — any open comments or threads already in the PR that the reviewer should be aware of. **Your initial read** — areas that warrant closer attention: logic that looks unusual, missing edge case handling, anything that doesn't align with team conventions, test coverage gaps, or any concerns the diff alone raises. Then stop and let the reviewer drive. ## Engage the Review The reviewer leads from here. Support whatever they need: - Drill into specific files, functions, or lines on request - Respond to or add to existing discussion threads via `gh pr comment [PR-number] --body "..."` via Bash — present any comment for confirmation before posting - Leave review comments on specific concerns via `gh pr review [PR-number] --comment --body "..."` via Bash — present the draft for confirmation before posting - Answer questions about what the code does or why it was written a particular way (inferred from the implementation and commit history) ## Actions Available When the reviewer is ready to act, offer the options that apply: **Approve** — if the reviewer confirms they are approving, run `gh pr review [PR-number] --approve` via Bash. Confirm once before acting. **Request changes** — run `gh pr review [PR-number] --request-changes --body "..."` via Bash with a summary of the reviewer's concerns. Present the draft body for review before posting. **Merge** — if the reviewer instructs a merge, check pipeline status first via `gh run list --branch [source-branch]` via Bash. Present the current state and get explicit confirmation before running `gh pr merge [PR-number]` via Bash. **Close without merging** — if the reviewer decides the PR should not proceed, update the PR with a closing comment via `gh pr comment [PR-number] --body "..."` and close it via `gh pr close [PR-number]` via Bash. Confirm before acting. No action happens without explicit reviewer instruction. No action happens without confirmation of what will be done before it is done. ## Thread Resolution After completing any round of review, instruct the reviewer: > "To mark addressed threads as resolved, open the PR in the GitHub UI and resolve each thread. GitHub does not support thread resolution via the CLI — this step must be done in the browser." This is expected behavior, not a workaround. Note which threads the reviewer indicated are addressed so they know which ones to resolve.
Related in Design
contribute
IncludedLocal-only OSS contribution command center. Auto-refreshes the user's in-flight PR and issue state on invoke so conversations start with full context — no need to brief Claude on what's in flight. Helps the user find issues to contribute to on GitHub, builds per-repo dossiers of what each upstream expects (CLA, DCO, branch convention, AI policy, draft-first, review bots, issue templates), runs deterministic gates before any external action so AI-assisted contributions don't reach maintainers as slop. State is markdown-only: candidate files at ~/.contribute-system/candidates/, repo dossiers at ~/.contribute-system/research/, append-only event log at ~/.contribute-system/log.jsonl. No database, no cloud calls. Use when the user asks about their PRs / issues / contributions, wants to find new work to take on, claim an issue, build/refresh a repo's dossier, or draft a Design Issue or PR. Trigger with "/contribute", "what's my PR status", "find a contribution", "claim issue X", "draft a Design Issue for Y", "refresh dossier for Z".
architectural-analysis
IncludedUser-triggered deep architectural analysis of a codebase or scoped subtree across eight modes — information architecture, data flow, integration points, UI surfaces, interaction patterns, data model, control flow, and failure modes. This skill should be used when the user asks to "diagram this codebase," "map the architecture," "show the data flow," "give me an ERD," "trace control flow," "find the integration points," "verify the layout pattern," "audit the UX architecture," or any similar request whose primary deliverable is mermaid diagrams plus cited reports under docs/architecture/. Dispatches haiku/sonnet sub-agents in parallel for per-mode exploration, then verifies every citation mechanically before any node lands in a diagram. Not for one-off prose explanations of code (use code-explanation) or for high-level system design from scratch (use system-design).
mcp
IncludedModel Context Protocol (MCP) server development and tool management. Languages: Python, TypeScript. Capabilities: build MCP servers, integrate external APIs, discover/execute MCP tools, manage multi-server configs, design agent-centric tools. Actions: create, build, integrate, discover, execute, configure MCP servers/tools. Keywords: MCP, Model Context Protocol, MCP server, MCP tool, stdio transport, SSE transport, tool discovery, resource provider, prompt template, external API integration, Gemini CLI MCP, Claude MCP, agent tools, tool execution, server config. Use when: building MCP servers, integrating external APIs as MCP tools, discovering available MCP tools, executing MCP capabilities, configuring multi-server setups, designing tools for AI agents.
react-native-skia
IncludedDesign, build, debug, and optimise high-polish animated graphics in React Native or Expo using @shopify/react-native-skia, Reanimated, and Gesture Handler. Use when the user wants canvas-driven UI, shaders, paths, rich text, image filters, sprite fields, Skottie, video frames, snapshots, web CanvasKit setup, or performance tuning for custom motion-heavy elements such as loaders, hero art, cards, charts, progress indicators, particle systems, or gesture-driven surfaces. Also use when the user asks for fluid, glow, glass, blob, parallax, 60fps/120fps, or GPU-friendly animated effects in React Native, even if they do not explicitly say "Skia". Do not use for ordinary form/layout work with standard views.
plaid
IncludedProduct Led AI Development — guides founders from idea to launched product. Six capabilities: Idea (discover a product idea), Validate (pressure-test the idea against fatal flaws, problem reality, competition, and 2-week MVP feasibility), Plan (vision intake + document generation), Design (translate image references into a design.md spec), Launch (go-to-market strategy), and Build (roadmap execution). Use when someone says "PLAID", "plaid idea", "help me find an idea", "product idea", "idea from my business", "idea from my expertise", "plaid validate", "validate my idea", "pressure-test", "is this idea good", "find fatal flaws", "validate the problem", "plan a product", "define my vision", "generate a PRD", "product strategy", "plaid design", "design from image", "translate image to design", "create design.md", "extract design tokens", "plaid launch", "go-to-market", "launch plan", "GTM strategy", "launch playbook", "plaid build", "build the app", "start building", or "execute the roadmap".
nextjs-framer-motion-animations
IncludedAdds production-safe Motion for React or Framer Motion animations to Next.js apps, including reveal, hover and tap micro-interactions, whileInView, stagger, AnimatePresence, layout and layoutId transitions, reorder, scroll-linked UI, and lightweight route-content transitions. Use when the user asks to add, refactor, or debug Motion or Framer Motion in App Router or Pages Router codebases, especially around server/client boundaries, reduced motion, LazyMotion, bundle size, hydration, or route transitions. Avoid for GSAP-style timelines, WebGL or 3D scenes, heavy scroll storytelling, or CSS-only effects unless Motion is explicitly requested.