kelly-financial-services-intel

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFEPROMPT_INJECTIONREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external news and regulatory sources to generate intelligence signals. Ingestion points: Data enters the system through monitored public URLs and central bank updates stored in Busabase. Boundary markers: The skill implements a review-first workflow where all signals and drafts remain in a 'needs_review' state until a human operator approves them. Capability inventory: The skill can read/write to Busabase and uses local scripts for decision handoff planning. Sanitization: The application interface employs HTML escaping in its rendering logic.
  • [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill includes utility scripts (execute_decisions.mjs, setup.mjs) for workspace management. These scripts use the busabase-sdk and Hono framework, which are standard and verified dependencies for this environment.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill handles financial intelligence and uses environment variables for secure credential management. It maintains data boundaries by isolating interactions to the user's specific Busabase workspace.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 10:36 AM
Security Audit — agent-trust-hub — kelly-financial-services-intel