alpha

Pass

Audited by Gen Agent Trust Hub on Mar 24, 2026

Risk Level: SAFECREDENTIALS_UNSAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill instructs the agent to solicit raw Binance API keys and secrets from the user via file uploads and store them in a local file named TOOLS.md.
  • [COMMAND_EXECUTION]: Documentation in references/authentication.md provides shell script examples that use curl, openssl, and date to perform cryptographic signing and execute network requests.
  • [DATA_EXFILTRATION]: The skill is designed to transmit sensitive authentication headers (X-MBX-APIKEY) and cryptographic signatures to Binance API endpoints.
  • [PROMPT_INJECTION]: The skill defines an attack surface for indirect prompt injection via user-provided parameters. Ingestion points: symbol, interval, fromId, and limit parameters in SKILL.md. Boundary markers: None identified to isolate user input from the request structure. Capability inventory: Shell command execution for API requests as defined in references/authentication.md. Sanitization: No explicit validation or escaping of user input is specified before interpolation into request strings.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 24, 2026, 01:34 PM