| |
| language: "en-US" |
| early_access: false |
| tone_instructions: "Be direct, concise, and evidence-led. Prioritize actionable findings over praise." |
|
|
| reviews: |
| profile: "assertive" |
| request_changes_workflow: false |
| high_level_summary: true |
| high_level_summary_in_walkthrough: true |
| review_status: true |
| review_details: true |
| commit_status: true |
| fail_commit_status: true |
| auto_review: |
| enabled: true |
| drafts: false |
| path_instructions: |
| - path: ".github/workflows/**" |
| instructions: | |
| Treat workflow changes as security-sensitive. Flag unpinned third-party actions, broad write permissions, persisted checkout credentials in write-token jobs, pull_request_target misuse, and untrusted GitHub context interpolated into shell commands. |
| - path: "{scripts,bin}/**" |
| instructions: | |
| Focus on command injection, unsafe subprocess usage, path traversal, SSRF, secret exposure, and missing tests for new CLI behavior. |
| - path: "skills/**/scripts/**" |
| instructions: | |
| Review generated or imported scripts as untrusted-input tooling. Flag RCE, path traversal, network fetches without validation, and writes outside the expected workspace. |
| - path: "{skills,commands,agents,rules}/**" |
| instructions: | |
| Focus on prompt-injection resilience, tool-permission scope, destructive action guards, and secret exfiltration risks. |
| - path: "{SECURITY.md,docs/security/**}" |
| instructions: | |
| Check that official distribution surfaces, disclosure guidance, and supply-chain rules stay accurate and do not endorse unofficial packages. |
| |
| chat: |
| auto_reply: true |
|
|
