broker-mcp-selector

Pass

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill directs the agent to fetch a live directory of broker capabilities from the author's GitHub repository (github.com/mphinance/awesome-broker-mcp). This process ensures that recommendations are based on verified, up-to-date information regarding broker safety postures.
  • [SAFE]: Implements proactive safety guidelines by instructing the agent to recommend the most restrictive permissions necessary for a task, such as choosing read-only servers for analysis-only workflows.
  • [DATA_EXFILTRATION]: Includes clear warnings about the risks of credential exposure when using third-party community servers and explicitly mentions the threat of indirect prompt injection affecting connected financial accounts, advising users to maintain separate environments for analysis and execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 29, 2026, 01:18 AM
Security Audit — agent-trust-hub — broker-mcp-selector