story-explore

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill operates in a restricted, read-only environment. It explicitly disables high-risk capabilities such as network access, file system modification, and shell execution.
  • [PROMPT_INJECTION]: The instructions include defensive measures to ignore attempts at bypassing safety filters or extracting system prompts. It specifically instructs the agent to disregard user-supplied IDs or unauthorized instructions embedded in natural language queries.
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing user input and story artifact data, but mitigates risk through capability restriction.
  • Ingestion points: Processes user queries and story artifact content (SKILL.md).
  • Boundary markers: Instructions explicitly mandate ignoring forged identifiers and natural language overrides.
  • Capability inventory: None; the skill is limited to read-only queries with no network or shell access.
  • Sanitization: Validates project and snapshot identifiers against the authorized execution environment.
  • [EXTERNAL_DOWNLOADS]: The skill metadata references a source repository on GitHub associated with the skill author. This reference is documented for attribution and does not involve runtime code execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 09:13 AM
Security Audit — agent-trust-hub — story-explore