tmdl
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because its primary purpose is to guide the agent in reading and modifying project files (.tmdl). If an attacker embeds malicious instructions within a project's TMDL metadata, descriptions, or comments, the agent might follow them while performing requested edits.
- Ingestion points: The agent is instructed to read and edit
.tmdlfiles within thedefinition/folder of a PBIP project. - Boundary markers: The skill does not provide specific instructions or markers to distinguish between data content and potential instructions within the TMDL files, although it provides strict semantic structure for the language itself.
- Capability inventory: The agent context typically includes file system read/write operations and potentially CLI tools for model validation.
- Sanitization: No specific sanitization or filtering of TMDL content is mandated before the agent processes it.
Audit Metadata