extract-design-system
Pass
Audited by Gen Agent Trust Hub on Jul 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
npxto execute theextract-design-systemCLI tool and install the Chromium browser driver. This is intended for the skill's primary function but involves running commands based on potentially untrusted parameters such as the target URL. - [EXTERNAL_DOWNLOADS]: The skill downloads the Chromium browser via the Playwright framework and fetches the
extract-design-systempackage from the npm registry. These downloads are associated with well-known services and the skill's own stated purpose. - [PROMPT_INJECTION]: The workflow introduces a surface for indirect prompt injection. In Step 3 of the
SKILL.mdworkflow, the agent is instructed to read and summarize the contents of.extract-design-system/normalized.json. Because this file is generated from a user-supplied public URL, a malicious website could craft content (e.g., in metadata or text) that, when extracted and reviewed, attempts to manipulate the agent's behavior or summary output.
Audit Metadata