The Agent Skills Directory

[SAFE]: The skill functions as a security auditing tool designed to protect users by identifying malicious code patterns in smart contracts. Its operations are consistent with its stated purpose of due diligence and risk assessment.
[COMMAND_EXECUTION]: The skill utilizes command-line tools including grep and a custom Python script (tools/token_scanner.py) to analyze source code files for known vulnerability signatures. This execution is performed locally and is necessary for the skill's analytical functions.
[PROMPT_INJECTION]: As an audit tool, the skill is designed to process external and potentially untrusted code. It mitigates indirect prompt injection risks by providing structured detection rules and strict criteria for evaluating the safety of external token logic.

meme-coin-audit