Skip to content

user3

user3 #5

name: Claude Security Review
on:
pull_request:
types: [opened, synchronize, ready_for_review, reopened]
# Optional: only run when security-sensitive areas change
# paths:
# - "src/auth/**"
# - "src/api/**"
# - "infra/**"
# - "config/**"
jobs:
security:
runs-on: ubuntu-latest
# Least privilege for PR review + required OIDC for the Claude GitHub App
permissions:
contents: read
pull-requests: write
id-token: write
steps:
- name: Checkout PR
uses: actions/checkout@v6
with:
fetch-depth: 1
- name: Claude security review
uses: anthropics/claude-code-action@v1
with:
# anthropic_api_key: ${{ secrets.ANTHROPIC_API_KEY }}
prompt: |
REPO: ${{ github.repository }}
PR NUMBER: ${{ github.event.pull_request.number }}
Perform a security-focused review. Prioritize:
- AuthN/AuthZ issues (BOLA/BFLA/BOPLA)
- Injection (SQLi/NoSQLi/Command)
- SSRF, deserialization, crypto misuse
- Sensitive data exposure, logging issues
- Security misconfig in code/config
For each finding, include:
Severity: CRITICAL/HIGH/MEDIUM/LOW
What/Where
Impact
Recommendation
Use inline comments for line-level issues and one top-level summary comment.
claude_args: |
--allowedTools "mcp__github_inline_comment__create_inline_comment,Bash(gh pr comment:*),Bash(gh pr diff:*),Bash(gh pr view:*)"