pr-code-review

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill explicitly instructs the agent to append staged/unstaged diffs and the full contents of untracked text files into the evidence/report, which can force the LLM to include secret values (API keys, passwords, tokens) verbatim in its output.

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

• Third-party content exposure detected (high risk: 0.75). Outsider-authored free text can enter the LLM context via PR/issue/discussion content fetched at runtime (e.g., gh pr view and “available PR text” used in step 4), which is then included in the evidence packet and dispatched to subagents.