stitched-full-page-capture

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill's core script scripts/stitch_full_page_capture.mjs executes system binaries ffmpeg and sips using node:child_process functions (spawn and execFile). These tools are used for cropping, stitching, and validating images, which is consistent with the skill's primary purpose and documented workflow.
  • [DYNAMIC_EXECUTION]: The script uses Playwright's page.evaluate() to execute JavaScript within the browser context to handle scrolling and calculate page dimensions. This is standard behavior for browser automation tools and is used here to ensure lazy-loaded content is rendered.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface for untrusted data as it navigates to external URLs provided in a manifest file using Playwright (page.goto).
  • Ingestion points: Reads pageUrl from a local manifest.json file and visits these URLs via a browser session.
  • Boundary markers: None identified for URL navigation; the tool relies on standard browser isolation.
  • Capability inventory: Uses playwright for browser control and ffmpeg/sips for image file processing.
  • Sanitization: Numerical inputs like quality and step are validated/casted, but the pageUrl is passed directly to the browser's navigation function.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 02:55 PM
Security Audit — agent-trust-hub — stitched-full-page-capture