infographic
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied topics and data to populate an HTML template. This is a standard data ingestion surface for infographic generation.
- Ingestion points: Data enters through the
/infographiccommand and optional user-provided data blocks. - Boundary markers: The instructions guide the agent to use specific HTML classes and data attributes to structure the output, which helps delimit data from instructional logic.
- Capability inventory: The skill involves writing local HTML files and executing a validation script that renders the content using Playwright.
- Sanitization: The provided
validate.pyscript serves as a security control by performing runtime checks for unauthorized external network requests and JavaScript errors during the rendering process, effectively mitigating exfiltration risks. - [DYNAMIC_EXECUTION]: The skill utilizes a validation script (
scripts/validate.py) that leverages Playwright to render generated HTML in a headless browser. This process is used exclusively for verifying the visual layout, accessibility, and security constraints of the infographic, and it is a core component of the skill's quality assurance workflow. - [EXTERNAL_DOWNLOADS]: When the optional handwritten style is enabled, the skill fetches specific fonts from Google Fonts. This external reference is limited to well-known, trusted domains (
fonts.googleapis.com,fonts.gstatic.com) and is explicitly permitted by the skill's validation logic. - [COMMAND_EXECUTION]: The skill instructions include steps for running a local Python validation script. This script is provided as part of the skill package and operates within the local file system scope to process the agent's output.
Audit Metadata