baoyu-infographic
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a template-driven design tool that operates using local reference files and standard agent tools. It does not perform suspicious network operations, execute arbitrary code, or attempt privilege escalation.
- [SAFE]: The skill follows security best practices by including explicit instructions to scan source material and strip out any credentials, tokens, or secrets before they are processed or stored in output files, mitigating data exposure risks.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it ingests untrusted user content for processing. This is documented with the following evidence chain:
- Ingestion points: User-supplied text, files, or URLs are loaded in Step 1.
- Boundary markers: The skill uses double-brace placeholders in its base prompt template and organizes data into a structured markdown format to isolate user content.
- Capability inventory: The skill's capabilities are restricted to file system operations (write_file/read_file) and image generation (image_generate).
- Sanitization: The instructions mandate the removal of secrets and require all data to be preserved verbatim, preventing the agent from interpreting instructions embedded within the source content.
Audit Metadata