msom-theory-development

Pass

Audited by Gen Agent Trust Hub on Aug 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of instructional content and checklists for academic writing. It does not perform any dangerous operations such as file system writes, network communication, or command execution.- [INDIRECT_PROMPT_INJECTION]: The skill has a data ingestion surface as it processes user-provided manuscript drafts. Ingestion points: User manuscripts and model ideas mentioned in the triggers. Boundary markers: Absent. Capability inventory: The skill uses only textual output and does not define access to tools, network, or filesystem. Sanitization: Absent. The risk is considered safe due to the lack of exploitable capabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 23, 2026, 06:57 AM
Security Audit — agent-trust-hub — msom-theory-development