The Agent Skills Directory

[COMMAND_EXECUTION]: The script dev_score.py uses subprocess.run() to execute gmgn-cli commands. This is standard and expected functionality for a skill designed to interface with a specific CLI tool for blockchain data analysis.
[EXTERNAL_DOWNLOADS]: The skill instructions provide guidance on installing the gmgn-cli tool via npm install -g gmgn-cli if it is not found. This is a common and legitimate practice for skills requiring specific local binaries to function.
[SAFE]: The skill uses a strict sanitization function (safe()) to clean all attacker-controlled strings (like token symbols or wallet addresses) before they are included in output, preventing potential prompt injection or terminal escape sequences. It also follows least-privilege principles by requesting only the specific CLI tools it needs for data gathering.

gmgn-dev-score