render-imessage-chat
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes ffmpeg, ffprobe, and playwright commands to perform its core video rendering functions. These executions are securely implemented using sanitized parameters and proper shell quoting in the stitch.sh script and record-chat.js.
- [EXTERNAL_DOWNLOADS]: Installation and smoke tests involve downloading standard browser binaries via Playwright (Chromium), which is necessary for the skill's recording functionality.
- [PROMPT_INJECTION]: While the skill ingests untrusted thread data that is rendered into a browser context, it mitigates indirect prompt injection risks through a robust escapeHTML sanitization function in generate.js.
- [DATA_EXPOSURE]: The skill performs legitimate local file system operations to read configuration and assets, with no evidence of unauthorized data exposure or exfiltration. Temporary files are managed securely using system-provided directory paths and cleaned up after execution.
Audit Metadata