openclaw-session-viewer
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted JSONL transcript data provided by the user or external agents. While this creates a surface for indirect prompt injection, the implementation uses comprehensive HTML escaping and sanitization (via
htmlEscapeandscriptJsonEscape) before rendering content in the browser, significantly mitigating XSS and injection risks. - [COMMAND_EXECUTION]: The skill provides an option to automatically open generated HTML files in the default browser. The implementation in
scripts/open-browser.tscorrectly avoids shell execution by passing arguments directly to the platform's opening command (open,explorer.exe, orxdg-open) and includes automated tests to verify that potentially malicious file paths do not lead to command injection.
Audit Metadata