rd-direction-finder
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill uses local shell and Python scripts (
scripts/render_report.py) to process retrieved data and generate HTML/Markdown reports. These operations are performed within the local environment on user-provided technical text for the purpose of report generation. - [EXTERNAL_DOWNLOADS]: The skill performs network searches via
mcp__patsnap_search__andweb.searchto gather research evidence. These operations target well-known academic and patent databases (CNKI, IEEE, Scopus, etc.) and are essential to the primary function of research direction finding. - [DATA_EXPOSURE_AND_EXFILTRATION]: The skill handles research descriptions (
problem_text) and retrieves technical metadata. It does not access sensitive system paths (~/.ssh, ~/.aws) or credentials. Data flow is directed into local artifact files (JSON/HTML/MD) within the@session/reports/directory. - [DYNAMIC_EXECUTION]: The skill utilizes
python3 << 'PYEOF'blocks to write configuration and data to local files. This is used for structured payload construction and does not incorporate untrusted external code for execution. - [PRIVILEGE_ESCALATION]: No privilege escalation commands (sudo, chmod 777) or persistence mechanisms were detected. The scripts operate with standard user permissions.
Audit Metadata