domain-model
Pass
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted data from the repository to inform its analysis and update project documentation.
- Ingestion points: The agent reads
CONTEXT.md,CONTEXT-MAP.md, files indocs/adr/, and source code withinsrc/as defined inSKILL.md. - Boundary markers: There are no specific instructions or delimiters defining boundaries to prevent the agent from obeying instructions that might be embedded within the project's documentation or source code.
- Capability inventory: The skill has the ability to explore the codebase and write to or update files like
CONTEXT.mdand ADRs. - Sanitization: The skill lacks explicit validation or sanitization steps for content retrieved from the codebase before using it to generate updates or influence the modeling session.
Audit Metadata