AI-powered PR review inside Claude Code. Analyzes code quality, security, performance, and best practices using Entelligence.AI.
Install and add to your Claude Code MCP config (~/.claude/.mcp.json):
pip install entelligence-pr-reviewer{
"mcpServers": {
"entelligence-pr-reviewer": {
"command": "entelligence-pr-reviewer"
}
}
}Restart Claude Code. On first use, you'll be prompted to enter your API key — no manual config needed.
/entelligence-review 42— Review PR #42 in your current repo/entelligence-review— Review your local uncommitted changes- Natural language — Just say "review PR #42" and Claude calls the tool automatically
pip install entelligence-pr-reviewerAdd to ~/.claude/.mcp.json (global) or .mcp.json (per-project):
{
"mcpServers": {
"entelligence-pr-reviewer": {
"command": "entelligence-pr-reviewer"
}
}
}On first use, the plugin will:
- Open your Entelligence dashboard in the browser
- Ask you to paste your API key
- Save it securely to
~/.entelligence/credentials.json
That's it. No config file editing required.
Already have a key? You can also set it via environment variable:
{
"mcpServers": {
"entelligence-pr-reviewer": {
"command": "entelligence-pr-reviewer",
"env": {
"ENTELLIGENCE_API_KEY": "ent-your-key-here"
}
}
}
}To get the /entelligence-review slash command, copy the skill to your project:
mkdir -p .claude/skills/entelligence-review
curl -o .claude/skills/entelligence-review/SKILL.md \
https://raw.githubusercontent.com/Entelligence-AI/entelligence-claude-code/main/.claude/skills/entelligence-review/SKILL.mdWithout the skill, the MCP tools still work — just say "review PR #42" in natural language.
/entelligence-review 42 # Review PR #42
/entelligence-review https://github.com/owner/repo/pull/42 # Review by URL
/entelligence-review # Review local changes
/entelligence-review staged # Review staged changes
Just tell Claude what you want:
- "Review PR #42"
- "Review this PR: https://github.com/owner/repo/pull/42"
- "Review my local changes"
- "Do a high-priority review of PR #100"
| Tool | Description |
|---|---|
review_pr |
Review a PR by number (async, handles large PRs) |
review_diff |
Review a raw diff string (sync, for small/local diffs) |
get_review_status |
Check status of a previously submitted review |
get_user_info |
Show your Entelligence account info |
save_api_key |
Save your API key (called automatically during setup) |
| Parameter | Values | Default | Description |
|---|---|---|---|
priority_level |
low, medium, high |
low |
Review depth |
mode |
concise, verbose |
concise |
Output verbosity |
Example: "Do a high-priority verbose review of PR #42"
| Environment Variable | Default | Description |
|---|---|---|
ENTELLIGENCE_API_KEY |
— | API key (or use interactive setup) |
ENTELLIGENCE_API_URL |
https://entelligence.ddbrief.com |
API base URL |
ENTELLIGENCE_DASHBOARD_URL |
https://app.entelligence.ai |
Dashboard URL |
You: "/entelligence-review 42"
→ Claude calls get_user_info (auth check)
→ Claude calls review_pr with repo + PR #42
→ MCP server submits async review job to Entelligence API
→ Polls until review is complete (~1-5 min)
→ Claude presents results with actionable fix suggestions
→ You can ask Claude to apply the fixes automatically
git clone https://github.com/Entelligence-AI/entelligence-claude-code.git
cd entelligence-claude-code
npm install
npm run build
# Point your .mcp.json to the local build for testing
# "args": ["/path/to/entelligence-claude-code/dist/index.js"]MIT