mantis-threat-model
Pass
Audited by Gen Agent Trust Hub on Jul 13, 2026
Risk Level: SAFE
Full Analysis
- [Ingestion of Architectural Data]: The skill reads from
workspace/kb/architecture.mdand theworkspace/kb/entities/directory to synthesize a threat model. This is a potential surface for indirect prompt injection if the source data contains adversarial content, though it is necessary for the skill's functionality. * Ingestion points:workspace/kb/architecture.mdandworkspace/kb/entities/*.md(SKILL.md). * Boundary markers: None present; the skill treats all KB content as valid input for analysis. * Capability inventory: The skill uses file-writing tools to generateworkspace/kb/THREAT_MODEL.md(SKILL.md). * Sanitization: None specified; the skill relies on the agent's synthesis process to format the output as Markdown. - [Workspace File Management]: The skill manages files within the
workspace/kb/directory. This automated maintenance of documentation helps keep security models up to date with architectural changes.
Audit Metadata