cowart-image-edit

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection by extracting instructions from text, labels, and annotations found within user-provided screenshots. Maliciously crafted text embedded in a screenshot could be used to influence the agent's behavior during the image generation or canvas modification process.
  • Ingestion points: The skill reads user-provided screenshots and extracts requirements from visible labels and notes (SKILL.md, Workflow Step 1 & 2).
  • Boundary markers: There are no defined boundary markers or instructions to treat extracted text as untrusted data.
  • Capability inventory: The agent has the ability to generate images, write files to the project directory, and modify the Cowart canvas state via the insert_cowart_image and save_cowart_canvas_state tools.
  • Sanitization: No sanitization or validation of the text extracted from images is described before it is passed to the generation prompt.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 03:18 PM
Security Audit — agent-trust-hub — cowart-image-edit