workflow-documenter
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external documentation files to perform automated edits on the project's source code and durable documentation.
- Ingestion points: The agent reads
docs/features/<feature-id>/spec.mdanddocs/features/<feature-id>/validator/validation-report.mdto identify necessary documentation updates and checklist items. - Boundary markers: The instructions lack explicit data delimiters or instructions to ignore embedded commands within the processed files, which could lead the agent to follow malicious instructions hidden in the specification documents.
- Capability inventory: The skill has the capability to modify
.ktfiles (adding KDoc) and various.mdfiles (ARCHITECTURE, README, DESIGN) using file-writing tools. - Sanitization: There is no evidence of sanitization, filtering, or validation of the content read from external files before it is used to modify the codebase.
Audit Metadata