industry-stock-tracker
Fail
Audited by Gen Agent Trust Hub on Jun 15, 2026
Risk Level: HIGHCREDENTIALS_UNSAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The script
scripts/generate_industry_stock_tracker_report.pycontains a hardcoded default API key (em_fjFqd4YB6Cqs52LF48XWbMDdLNq6MyNg) within theEM_API_KEYvariable assignment. - [COMMAND_EXECUTION]: The
SKILL.mdfile defines a workflow that executes a Python script via shell command, interpolating user-provided input directly into the command arguments (--query "{{query}}"). - [EXTERNAL_DOWNLOADS]: The skill communicates with
ai-saas.eastmoney.comto fetch tracking reports and downloads file attachments in PDF and DOCX formats to the local filesystem. - [DATA_EXFILTRATION]: User queries provided to the agent are transmitted to the external domain
ai-saas.eastmoney.comto facilitate report generation. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface.
- Ingestion points: External data is ingested from the remote API via the
titleandcontentfields inscripts/generate_industry_stock_tracker_report.py. - Boundary markers: No delimiters or instructions are used to distinguish external report content from the agent's internal logic.
- Capability inventory: The script performs file system writes (
Path.write_bytes) and network requests (urllib.request). - Sanitization: The
_clean_report_textfunction performs basic formatting cleanup but does not sanitize or filter the content for malicious instructions or command sequences.
Recommendations
- AI detected serious security threats
Audit Metadata