wiki-qa

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt requires collecting and embedding REPO_URL (and user-provided repository URLs) verbatim into linked citations/outputs (and runs git commands that can reveal remote URLs), so any credentials or tokens embedded in those URLs would be output directly and exfiltrated.

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md "Source Repository Resolution" requires the agent to accept a repository URL (e.g., GitHub) and then search/read those source files (steps: "store as REPO_URL", "Search the codebase", "Read those files to gather evidence"), so arbitrary public/untrusted repositories can be ingested and directly influence the agent's outputs.