creative-status
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from local project files, creating a surface for indirect instructions to influence agent behavior.
- Ingestion points: Reads project-specific data from
.creative/config.yaml,.creative/status.json, and various content files (e.g., drafts, planning layers) defined in the domain configuration. - Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions when interpolating project content into the agent's context.
- Capability inventory: The skill performs file globbing, file reading, and logic-based statistics calculation. It does not possess network access, file-write permissions, or arbitrary code execution capabilities.
- Sanitization: No evidence of sanitization or content filtering for the ingested project data is present.
- [COMMAND_EXECUTION]: The skill uses local shell commands for environment verification.
- Evidence: Executes a directory check (
if [ ! -d ".creative" ]) to ensure the agent is operating within a valid project folder before proceeding. This is a benign environment check.
Audit Metadata