docs-manage

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt explicitly supports and exemplifies embedding Authorization headers (e.g., --header "Authorization: Bearer tok_xxx") and custom headers on the command line, which would require inserting secret tokens verbatim into generated commands/outputs and therefore poses an exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md "scrape" command explicitly downloads and indexes documentation from arbitrary URLs (e.g., npx ... scrape react https://react.dev/...), so the agent fetches and ingests open/public third-party web content that could contain instructions influencing later behavior.