birdeye-plugin

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads its runtime components and version metadata from the official GitHub repository of the OKX organization. These downloads include integrity verification using SHA256 checksums before installation.
  • [DATA_EXFILTRATION]: Implements robust protection against credential exposure by isolating Solana private key handling within a separate host process and providing explicit instructions to deny the agent direct read access to the key file.
  • [COMMAND_EXECUTION]: Uses standard shell commands for the specific purpose of skill installation, environment verification, and update checking. The commands do not exhibit signs of arbitrary execution or malicious intent.
  • [PROMPT_INJECTION]: Effectively mitigates indirect prompt injection risks by sanitizing and filtering all API response data through a whitelist of safe fields before surfacing information to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 03:28 AM
Security Audit — agent-trust-hub — birdeye-plugin