hence-search

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt explicitly shows and recommends passing an API key directly on the command line (python scripts/auth.py <api-key>), which encourages embedding secrets verbatim in commands and therefore creates an exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.70). The skill explicitly fetches and ingests user-generated projects from the open Hence gallery via public API endpoints (see SKILL.md "Search the Hence gallery" and references/api.md plus runtime calls in scripts/search.py and scripts/fetch_metadata.py), and the agent reads titles/pitches/metadata and uses them to drive recommendations, memory ("save_memory") and follow-up actions, so untrusted third-party content could indirectly influence behavior.