vp-review

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt requires the reviewer to read artifacts and "quote the problematic text" as evidence (and to read source code/files when grounding), which can force verbatim reproduction of any API keys, tokens, or passwords found in those artifacts, creating an exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.80). This skill's Phase 2: Ground workflow explicitly requires the agent to "do targeted web searches," "read official documentation," and "check OSS repos" (see "When the artifact involves external technologies, APIs, or frameworks" in SKILL.md), meaning it will fetch and interpret public third‑party content that can influence decisions and tool use, creating a clear indirect prompt‑injection exposure.