hackernews-frontpage
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and evaluate content from Hacker News via the SandBase API tools, creating a potential surface for indirect prompt injection from external web content. \n
- Ingestion points: External data retrieved via sandbase_run and sandbase_run_get (SKILL.md, references/sandbase-api-map.md). \n
- Boundary markers: The instructions do not define specific technical delimiters for external data but enforce strict process boundaries, requiring the agent to 'Return source-linked findings' and 'distinguish observed metadata from qualitative recommendations.' \n
- Capability inventory: The skill uses network-capable vendor tools for data retrieval and inference but does not include capabilities for file system modification or local code execution. \n
- Sanitization: Relies on qualitative verification and cross-checking against primary sources rather than programmatic sanitization.
Audit Metadata