dflow-kalshi-market-scanner

Pass

Audited by Gen Agent Trust Hub on Apr 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill communicates exclusively with DFlow's official Metadata API endpoints and does not perform any sensitive data exfiltration or unauthorized file access.
  • [SAFE]: Credential management follows security best practices; the skill instructs the agent to prompt the user for an API key rather than relying on insecure defaults or hardcoded values.
  • [SAFE]: Script generation is limited to fetching public market data and providing a bridge to the official dflow CLI, which is consistent with the skill's stated purpose.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection via market metadata. Ingestion points: Market titles and subtitles from pond.dflow.net. Boundary markers: Absent. Capability inventory: Script generation with dflow CLI shell-outs. Sanitization: Absent. Note: This surface is necessary for the skill's primary purpose and is considered safe in this context.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 23, 2026, 04:01 AM