pancakeswap

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill's reference scripts (e.g., farm-apr.py, fetch-farms.py) dynamically install the requests library from the official PyPI registry. Instructions also guide the installation of well-known Node.js packages from the NPM registry. These are standard operations for the skill's primary DeFi functions.
  • [COMMAND_EXECUTION]: The skill uses subprocess.check_call and os.system within its local reference scripts to perform static package installation (e.g., pip install requests -q). These operations are safe as they do not incorporate untrusted or dynamic user input into the command string.
  • [INDIRECT_PROMPT_INJECTION]: The skill interacts with external DeFi APIs such as DexScreener, CoinGecko, and DefiLlama to retrieve token metadata and pool statistics. To mitigate the risk of processing malicious data from these untrusted sources, the instructions mandate strict regex validation for all blockchain addresses and enforce the use of single-quoting for all shell variables to prevent command injection.
  • [SAFE]: Telemetry and initialization pings are directed to pancakeswap.ai, the official domain for the service. The data transmitted (e.g., OS, architecture, timestamp, and system identifier) is standard for telemetry and session initialization and does not include sensitive user secrets.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 04:28 AM
Security Audit — agent-trust-hub — pancakeswap