The Agent Skills Directory

[SAFE]: The skill provides legitimate technical documentation and integration examples for the official Uniswap Trading API and Universal Router.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests untrusted user data to generate code and API requests, it implements effective security mitigations:
Ingestion points: User-provided token symbols, amounts, and wallet addresses are processed during the generation of scripts and API calls.
Boundary markers: The skill includes an 'Input Validation Rules' section that enforces strict schemas for external data.
Capability inventory: The skill utilizes WebFetch and Bash for network operations and Write/Edit for file management.
Sanitization: Implements regex-based validation for Ethereum addresses and numeric amounts, and explicitly rejects shell metacharacters (e.g., ;, |, $, backticks) to prevent command injection.
[SAFE]: Follows security best practices for secret management by instructing the agent to use environment variables rather than hardcoding API keys in generated code.
[SAFE]: Includes a mandatory user confirmation step via AskUserQuestion before any transaction that spends gas or transfers tokens is submitted, ensuring human-in-the-loop oversight for financial actions.
[EXTERNAL_DOWNLOADS]: All external libraries (e.g., @uniswap/universal-router-sdk, viem) and API endpoints (trade-api.gateway.uniswap.org) originate from well-known and reputable sources in the decentralized finance ecosystem.

swap-integration