scriptcat-userscript
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Node.js scripts, specifically
scripts/check.mjsfor static analysis andscripts/stub.mjsfor generating development stubs. It also utilizes thesctlCLI tool to interact with the ScriptCat browser extension and the user's browser.\n- [DYNAMIC_EXECUTION]: The skill facilitates the execution of JavaScript code within a browser page context using thesctl page evalcommand for the purpose of scouting page elements and verifying userscript behavior.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests and analyzes userscripts and reads external documentation, representing a theoretical attack surface for indirect prompt injection. This is categorized as low-risk given its primary function as a developer tool.\n - Ingestion points: Reading local userscript files and remote API type definitions (scriptcat.d.ts).\n
- Boundary markers: The instructions do not define specific delimiters for separating analyzed script content from the agent's instructions.\n
- Capability inventory: The agent can execute shell commands (node, sctl), write local files (scripts/stub.mjs), and evaluate code in a browser context.\n
- Sanitization: No explicit sanitization or filtering of analyzed script content is specified.\n- [EXTERNAL_DOWNLOADS]: The skill fetches technical reference data, such as type definitions and code examples, from the official ScriptCat GitHub repository.
Audit Metadata