n8n Workflow: Secure GitHub PR Reviews with Claude, GPT-4 & Jira

Automatically review every GitHub pull request—securely

This n8n workflow automatically analyzes new GitHub pull requests using Claude (Anthropic) and GPT-4 (OpenAI), checks dependencies with Snyk, posts results back to GitHub, creates Jira issues for critical findings, logs everything to PostgreSQL, and notifies your team in Slack.

What this workflow does

• Triggers on new GitHub pull requests in your configured repository.
• Fetches PR metadata and diffs (changed files and patch content) via the GitHub API.
• Runs parallel analysis on parsed per-file diffs:
  • Uses Anthropic Claude for code review plus security and performance analysis.
  • Uses OpenAI GPT-4 for test coverage feedback.
• Performs targeted checks for specific file types, including:
  • Code complexity metrics
  • Secret detection
  • NPM dependency vulnerability scanning via the Snyk API
• Posts formatted GitHub review comments and requests changes on high-risk pull requests.
• Creates Jira tickets for findings flagged at the configured critical severity.
• Updates GitHub status/checks and writes review/developer statistics to PostgreSQL, while sending a Slack notification with an aggregated summary and risk score.

Use cases

• SaaS teams that want consistent, automated secure PR reviews before code merges
• Engineering orgs using Jira to track critical security/performance issues from PRs
• Teams standardizing dependency risk checks with Snyk and surfacing results directly in GitHub

Technical details

• Integrations: GitHub, Anthropic Claude, OpenAI GPT-4, Snyk, Jira, Slack, PostgreSQL
• Workflow nodes mentioned: if, set, code, jira, merge, slack
• Configuration: connect GitHub OAuth2, ensure GitHub API credentials for PR review actions, and set Anthropic credentials (plus the other connected services)