@554/chart

Pass

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes Playwright to automate a headless Chromium browser for rendering and capturing high-fidelity screenshots of generated charts in scripts/build_chart.py.
  • [EXTERNAL_DOWNLOADS]: The chart templates load the Apache ECharts library from the JSDelivr CDN, which is a well-known and recognized service for distributing front-end dependencies.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it interpolates data and text provided by the agent directly into HTML and JavaScript templates. Ingestion points: Data enters the process through the title, subtitle, and replacements parameters in scripts/build_chart.py. Boundary markers: No explicit markers or warnings are used to delimit untrusted data within the templates. Capability inventory: The skill can write files to the workspace, execute browser operations via Playwright, and run a local HTTP server. Sanitization: There is no evidence of HTML escaping or input validation on the data interpolated into the chart templates, creating an injection surface.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 15, 2026, 11:26 AM