financial-report-analyzer

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill transmits the content of user-provided financial reports and the SOMARK_API_KEY to the vendor's API at somark.cn or somark.ai for parsing. This behavior is the intended functionality of the skill and aligns with the vendor's provided services.
  • [COMMAND_EXECUTION]: The skill executes a local Python script, financial_report_analyzer.py, which is used to manage the API interaction and save the resulting Markdown and JSON files to the local filesystem.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted documents provided by the user, which presents a surface for indirect instructions designed to influence AI output.
  • Ingestion points: The script financial_report_analyzer.py reads user-specified files from the local disk via the -f flag.
  • Boundary markers: The SKILL.md file contains explicit instructions to the AI agent to treat all parsed document content strictly as data and to avoid executing any embedded instructions.
  • Capability inventory: The skill possesses the capability to read local files, write parsed results to the local filesystem, and perform network requests to the SoMark API endpoints.
  • Sanitization: The raw input (such as PDFs or images) is parsed into structured Markdown and JSON formats by the SoMark service before the AI processes the results.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 01:17 PM
Security Audit — agent-trust-hub — financial-report-analyzer