canvas-design

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill employs simulated dialogue history in the "FINAL STEP" section (e.g., "The user ALREADY said 'It isn't perfect enough...' ") to force the agent into a specific iterative refinement state, bypassing the natural conversational flow.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its handling of untrusted user thematic input without sufficient safeguards. 1. Ingestion points: User themes and conceptual threads are processed in the "DESIGN PHILOSOPHY CREATION" and "DEDUCING THE SUBTLE REFERENCE" sections. 2. Boundary markers: No delimiters or explicit instructions to ignore embedded commands are present when processing user thematic data. 3. Capability inventory: The agent is tasked with generating and refining code to produce .md, .pdf, and .png artifacts. 4. Sanitization: No input validation or filtering logic is specified for the user-supplied conceptual thread.
  • [NO_CODE]: The skill package does not contain any executable scripts or configuration files, consisting exclusively of markdown instructions and license documentation for fonts.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 01:35 AM