repo-report-builder

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly asks whether sensitive values should be "fully included for engineering audit" and instructs the agent to scan repos and include evidence (file/line), which would require outputting secret values verbatim if the user opts in, creating a clear exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.75). The required workflow ingests repository source code (including README/docs/issues/PR text authored by outsiders) via the repo-clone/fetch + “Repository Analysis Workflow” steps, and that code/docs text can become LLM-readable prose in the agent’s context for report generation.