challenge

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The protocol explicitly requires the agent to intake "claim + evidence verbatim" and to emit the "Claim challenged: " field, which forces the LLM to reproduce any secret values present in the input verbatim in its output.

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md specifies that at runtime the challenger protocol performs a counter-example hunt by searching/fetching outsider-selected content via Context7 -> Exa -> fuse-browser (browser_fetch/browser_serp_batch/browser_crawl) based on a user-provided claim and evidence intake, making free text from the user indirectly ingested through the browsing/search results path.