filings-edgar-ingest

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes financial filings from external sources (SEC EDGAR) which presents a surface for indirect prompt injection. Malicious instructions embedded within a filing could attempt to influence the agent's behavior when the data is subsequently retrieved.\n
  • Ingestion Points: download_filing fetches content from www.sec.gov.\n
  • Boundary Markers: There are no explicit delimiters or directives used to instruct the agent to ignore instructions within the retrieved chunks.\n
  • Capability Inventory: The skill has database write capabilities via psycopg2 and network access via requests.\n
  • Sanitization: The extract_section_text function uses lxml to remove HTML tags, but it does not validate or sanitize the semantic content of the text.\n- [EXTERNAL_DOWNLOADS]: Fetches company metadata and official filing documents from the SEC EDGAR API and document archive.\n- [COMMAND_EXECUTION]: Documentation provides an example for scheduling the ingestion workflow using a crontab entry for automated daily updates.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 10:43 AM
Security Audit — agent-trust-hub — filings-edgar-ingest