skills/iqrow/skills/teardown/Gen Agent Trust Hub

teardown

Pass

Audited by Gen Agent Trust Hub on Nov 3, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill performs network requests to verify source citations in teardowns (kit/verify.mjs) and to fetch remote teardown configuration files (kit/teardown.mjs). This is a core part of its verification functionality.
  • [COMMAND_EXECUTION]: The skill automates headless Chrome to validate the rendered output and generate preview images (kit/chrome.mjs). This is implemented through standard DevTools protocol interaction with the browser process.
  • [DYNAMIC_EXECUTION]: Headless Chrome is used to evaluate JavaScript within the browser environment to interact with the rendered teardown for validation and screenshot generation.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes structured subject data and external web content, which constitutes a potential ingestion surface for indirect prompt injection.
  • Ingestion points: Data is loaded from local teardown folders or remote JSON files in kit/teardown.mjs. Source verification involves fetching external URLs defined in the metadata in kit/verify.mjs.
  • Boundary markers: Absent. The skill treats ingested subject data as trusted structured content.
  • Capability inventory: Includes network access via fetch, local file system reads scoped to the teardown folder, and headless browser automation.
  • Sanitization: The skill normalizes fetched HTML content into plain text for citation verification, which helps isolate the execution environment from external scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 3, 2026, 09:08 AM