Skills
skills/agiprolabs/claude-trading-skills/mev-analysis/Gen Agent Trust Hub

mev-analysis

Pass

Audited by Gen Agent Trust Hub on Mar 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches market data, including token liquidity and volume, from DexScreener and CoinGecko public APIs.
  • [EXTERNAL_DOWNLOADS]: Communicates with Solana mainnet RPC nodes and Helius infrastructure for transaction parsing and analysis.
  • [EXTERNAL_DOWNLOADS]: Interacts with the Jito Block Engine for transaction bundle submission and status tracking.
  • [COMMAND_EXECUTION]: Provides instructions for installing the standard httpx Python library via the uv package manager.
  • [PROMPT_INJECTION]: Vulnerable to indirect prompt injection through the ingestion of untrusted data from blockchain RPCs and market data providers.
  • [PROMPT_INJECTION]: Ingestion points: fetch_token_pool_data in scripts/mev_risk_estimator.py and fetch_transaction in scripts/sandwich_detector.py.
  • [PROMPT_INJECTION]: Boundary markers: Absent.
  • [PROMPT_INJECTION]: Capability inventory: External network communication via the httpx library.
  • [PROMPT_INJECTION]: Sanitization: Standard JSON parsing is used without additional validation of content fields before processing or display.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 21, 2026, 02:36 PM