The Agent Skills Directory

[SAFE]: The skill integrates with official Uniswap infrastructure (https://liquidity.api.uniswap.org). This domain is consistent with the 'uniswap' author context and recognized as a legitimate vendor resource.
[PROMPT_INJECTION]: The instructions include robust input validation rules. It explicitly instructs the agent to reject inputs containing shell metacharacters and enforces strict regex patterns for Ethereum addresses, chain IDs, and token amounts.
[CREDENTIALS_UNSAFE]: The skill correctly handles API key management. It explicitly warns against hardcoding keys and instructs users to load them from environment variables.
[DATA_EXFILTRATION]: No evidence of unauthorized data access or exfiltration. The skill follows a non-custodial architecture where the API only builds transactions, while the user's application handles signing and broadcasting locally.
[COMMAND_EXECUTION]: The skill utilizes legitimate tools for developer tasks (Bash/npm/curl) but restricts their usage to documentation and API interaction contexts. It mandates an AskUserQuestion confirmation flow before any blockchain transaction is executed, preventing autonomous fund movement.

lp-integration