to-spec-beats
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external article files, which represents a potential attack surface for indirect prompt injection.
- Ingestion points: The skill reads
content/ARTICLE.mdandcontent/TOPIC.mdas primary inputs. - Boundary markers: The instructions require source files to include a
status: approvedfield and anapproved_digestin their metadata. - Capability inventory: The skill's capabilities are restricted to writing the
BEATS.mdfile and invoking theartifact-previewsub-skill. - Sanitization: The skill mitigates risks by requiring the agent to recompute and verify the SHA-256 hash of the input files against the stored digest before proceeding.
- [SAFE]: No evidence of prompt injection, unauthorized command execution, or network exfiltration was found. The skill's reliance on integrity verification for file processing is a proactive security measure.
Audit Metadata