dreamina-3d-resume
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes job state information from external files, which creates a potential surface for indirect prompt injection if the files are manipulated by an attacker.
- Ingestion points: In
SKILL.md, the workflow begins by callingJobLedger.read()on a user-specified job file to determine the next action. - Boundary markers: The instructions do not define specific delimiters or "ignore instructions" guards for the content read from the ledger file.
- Capability inventory: The skill possesses the ability to perform network requests (via
McpDesignClient), query account status, and read/write job state files. - Sanitization: While the skill requires re-validating preview receipts (as noted in the
PreviewSpecifiedstate mapping), it lacks explicit sanitization or strict schema validation for the job ledger's internal structure before mapping it to actions.
Audit Metadata