github-cli

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt includes and encourages insecure patterns (e.g., passing secrets inline like gh secret set API_KEY --body "sk-...") which can lead an agent to embed secret values verbatim in commands, even though it doesn't strictly force the LLM to request or print actual secrets.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow uses gh to view issue/PR details and logs (e.g., gh issue view ... --json body,comments --jq ... and gh run view ... --log), which ingests outsider-authored free text such as issue/PR bodies and comments written by other users into the LLM context via the command output.