sustainability-summary

Warn

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: MEDIUMPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION] (MEDIUM): Indirect prompt injection surface via untrusted RSS ingestion.\n
  • Ingestion points: Reads from sustainability_rss.db tables (entries, entry_content) containing external web content.\n
  • Boundary markers: None identified; untrusted content is not delimited from agent instructions.\n
  • Capability inventory: Agent performs summarization and evidence anchoring based on retrieved text.\n
  • Sanitization: No evidence of validation or sanitization of external data before context construction.\n- [COMMAND_EXECUTION] (LOW): Local command execution of a script provided in the skill package.\n
  • Evidence: SKILL.md triggers python3 scripts/time_report.py. The script itself was not provided for analysis.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 16, 2026, 10:56 AM