story-summarizer
Pass
Audited by Gen Agent Trust Hub on Feb 19, 2026
Risk Level: SAFE
Full Analysis
- [SAFE] (SAFE): No malicious patterns detected. The skill is purely informational and uses the 'Read' tool for text analysis.
- [Indirect Prompt Injection] (SAFE): The skill processes untrusted story text. Although it lacks explicit boundary markers to separate instructions from data, its capabilities are restricted to reading, posing no risk of data exfiltration or system compromise.
- Ingestion points: User-provided story text defined in the input requirements section of SKILL.md.
- Boundary markers: Absent; there are no specific markers to distinguish story content from potential embedded instructions.
- Capability inventory: Only the 'Read' tool is authorized, with no access to network, filesystem writes, or shell execution.
- Sanitization: No input sanitization or validation logic is present.
Audit Metadata