story-research

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill has a defined attack surface for indirect prompt injection as it is designed to ingest data from untrusted external URLs and web search results.
  • Ingestion points: Processes external URLs, attachments, and web search snippets provided via user input or automated retrieval (documented in SKILL.md).
  • Boundary markers: The skill instructions mandate that all external content be tagged as untrusted_source and explicitly states that instructions found within external webpages must not be executed.
  • Capability inventory: The skill utilizes tool-based web searching and fetching, delegating detailed analysis to a story-researcher subagent.
  • Sanitization: Includes specific prohibitions against accessing internal networks, cloud metadata, and local files, and strictly forbids the exfiltration of user credentials, cookies, or full manuscripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 09:13 AM
Security Audit — agent-trust-hub — story-research