enhance-prompt
Pass
Audited by Gen Agent Trust Hub on Feb 21, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill is susceptible to indirect prompt injection via the processing of the
DESIGN.mdfile. An attacker who can influence the contents of this file could potentially inject instructions that override the agent's behavior during the enhancement process. - Ingestion points: The skill reads the contents of
DESIGN.mdin Step 2 of the pipeline. - Boundary markers: There are no explicit delimiters or "ignore instructions" warnings defined when the content of
DESIGN.mdis extracted and formatted into the "DESIGN SYSTEM (REQUIRED)" block. - Capability inventory: The skill is granted
ReadandWritepermissions, allowing it to modify project files based on the (potentially poisoned) enhanced prompt. - Sanitization: No sanitization or validation is performed on the data read from the local file system before it is interpolated into the output prompt.
- [Trusted Source] (SAFE): The installation instructions use a repository hosted under a trusted organization (Google), and the external documentation links to a verified trusted domain.
Audit Metadata