create-claude-reviewer
Create a Claude Code Review GitHub Action workflow for PRs. Use when the user asks to "set up claude review", "add PR review", "create code review action", "claude reviewer", "set up automated review", or wants automated PR reviews with Linear or GitHub comments.
What this skill does
# Create Claude Code Review GitHub Action
You are creating a `claude-code-review.yml` GitHub Action workflow that automatically reviews PRs using Claude. This workflow uses `anthropics/claude-code-action@v1`.
There are two modes:
- **Linear mode** — Creates structured issues in Linear under the parent ticket (no PR comments)
- **GitHub mode** — Writes inline review comments directly on the PR using `gh`
## Context Awareness
This skill can be invoked at any point. Check conversation context — the user might reference repos already discussed, a workspace they just set up, or specific sub-repos. Use `$ARGUMENTS` if provided, otherwise infer from context.
**Important**: This workflow gets created in each working repo individually (each sub-repo in a meta-repo setup, NOT the meta-repo root). Each repo needs its own workflow because review rules and repo-specific violations differ per project.
---
## Step 1: Ask Clarifying Questions
Use `AskUserQuestion` to gather what you need. Skip anything obvious from context.
**Always ask:**
- **Linear or GitHub?** — "Should review findings be posted as Linear issues or as inline GitHub PR comments?"
**Ask if not obvious from context:**
- **Which repos?** — If in a meta-repo, which sub-repos should get the workflow? (default: all)
- **Target branch** — What branch do PRs target? (e.g., `develop`, `main`)
- **Model** — Which Claude model? (default: `claude-opus-4-6`)
**Ask only for Linear mode:**
- **Branch naming convention** — How are branches named? Need the pattern to extract ticket IDs (e.g., `wha-XXXX-description` extracts `WHA-XXXX`)
- **Linear team UUID** — What Linear team should issues be created in?
- **Linear label UUIDs** — Do they have label UUIDs for critical/warning severity? (optional, can skip)
- **Linear state preferences** — What state should review issues be created in? (default: "Todo")
---
## Step 2: Research Each Repo
For each repo that will get the workflow, read:
- `CLAUDE.md` — for repo-specific rules and violations to watch for
- `package.json` / `pyproject.toml` / `requirements.txt` — to understand the tech stack
- Existing `.github/workflows/` — check if a claude review workflow already exists
Extract from each repo:
1. **Tech stack** — language, framework (affects what review rules make sense)
2. **Repo-specific rules** — from CLAUDE.md, things the reviewer should flag as violations
3. **Existing workflow** — if one exists, ask the user if they want to replace or skip
---
## Step 3: Generate the Workflow
Create `.github/workflows/claude-code-review.yml` in each target repo.
### GitHub Mode Template
GitHub mode is simpler — no MCP config needed. Claude reviews the PR and posts inline comments using the built-in PR comment functionality.
```yaml
name: Claude Code Review
on:
pull_request:
types: [opened, synchronize]
branches:
- [TARGET_BRANCH]
jobs:
claude-review:
runs-on: ubuntu-latest
permissions:
contents: read
pull-requests: write
issues: read
id-token: write
steps:
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 1
- name: Run Claude Code Review
uses: anthropics/claude-code-action@v1
with:
claude_code_oauth_token: ${{ secrets.CLAUDE_CODE_OAUTH_TOKEN }}
github_token: ${{ secrets.GITHUB_TOKEN }}
use_sticky_comment: true
prompt: |
## PR Code Review
You are reviewing PR #${{ github.event.pull_request.number }} on branch `${{ github.head_ref }}`.
### Review Focus - REAL ISSUES ONLY
Review the PR diff for CONCRETE, SIGNIFICANT issues only:
- **Security vulnerabilities** (injection, auth bypass, data exposure, XSS, CSRF)
- **Performance problems** (N+1 queries, memory leaks, unbounded loops)
- **Actual bugs** that will cause runtime errors or incorrect behavior
[REPO_SPECIFIC_RULES]
**DO NOT REPORT:**
- Edge cases or rare scenarios unlikely to occur
- Style preferences or minor code organization suggestions
- Theoretical issues that "could" happen but probably won't
- Nitpicks or "nice to have" improvements
- Issues without clear, concrete impact
### Output Format
For each finding, use `gh` to post an inline review comment on the specific file and line:
```bash
gh api repos/${{ github.repository }}/pulls/${{ github.event.pull_request.number }}/comments \
--method POST \
-f body="**[Critical]** or **[Warning]**: Description of the issue and its concrete impact" \
-f commit_id="$COMMIT_SHA" \
-f path="path/to/file" \
-F line=LINE_NUMBER
```
Use `gh pr diff ${{ github.event.pull_request.number }}` to get the diff and identify exact file paths and line numbers.
If no significant issues are found, that's a GOOD outcome. Post a single summary comment saying the review is clean.
**ONLY report genuine, concrete concerns with clear impact.**
claude_args: |
--model [MODEL]
--allowedTools Bash(gh pr diff:*),Bash(gh pr view:*),Bash(gh api:*)
```
### Linear Mode Template
Linear mode creates structured issues in Linear, organized under the parent ticket. Requires Linear MCP config.
```yaml
name: Claude Code Review
on:
pull_request:
types: [opened, synchronize]
branches:
- [TARGET_BRANCH]
jobs:
claude-review:
runs-on: ubuntu-latest
permissions:
contents: read
pull-requests: write
issues: read
id-token: write
steps:
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 1
- name: Create MCP Config
env:
LINEAR_TOKEN: ${{ secrets.[LINEAR_SECRET_NAME] }}
run: |
cat > /tmp/mcp-config.json << EOF
{
"mcpServers": {
"linear": {
"command": "npx",
"args": [
"-y",
"mcp-remote@latest",
"https://mcp.linear.app/mcp",
"--header",
"Authorization:Bearer ${LINEAR_TOKEN}"
]
}
}
}
EOF
- name: Run Claude Code Review
uses: anthropics/claude-code-action@v1
with:
claude_code_oauth_token: ${{ secrets.CLAUDE_CODE_OAUTH_TOKEN }}
github_token: ${{ secrets.GITHUB_TOKEN }}
use_sticky_comment: true
prompt: |
## PR Review - Create Linear Issues
You are reviewing PR #${{ github.event.pull_request.number }} on branch `${{ github.head_ref }}`.
### Step 1: Extract Linear Ticket ID
Extract the ticket ID from the branch name. Branch format is `[BRANCH_PATTERN]`.
The ticket ID is the `[PREFIX]-XXXX` part.
### Step 2: Review the PR - FOCUS ON REAL ISSUES ONLY
Review the PR diff for CONCRETE, SIGNIFICANT issues only:
- **Security vulnerabilities** (injection, auth bypass, data exposure, XSS, CSRF)
- **Performance problems** (N+1 queries, memory leaks, unbounded loops)
- **Actual bugs** that will cause runtime errors or incorrect behavior
[REPO_SPECIFIC_RULES]
**DO NOT REPORT:**
- Edge cases or rare scenarios unlikely to occur
- Style preferences or minor code organization suggestions
- Theoretical issues that "could" happen but probably won't
- Nitpicks or "nice to have" improvements
- Issues without clear, concrete impact
### Step 3: Check Existing Issues Before Creating New Ones
1. Look up the parent ticket using `mcp__lineaRelated in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.