The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the polygon-agent CLI tool and instructs the agent to execute various shell commands for wallet setup, financial transactions, and API interactions.
[EXTERNAL_DOWNLOADS]: The skill instructs the installation of the @polygonlabs/agent-cli global NPM package. It also dynamically references and loads additional instruction sub-skills from https://agentconnect.polygon.technology/.
[DATA_EXFILTRATION]: The x402-pay command is used to send data, including user code snippets, to an external service at https://x402-api.onrender.com/api/code-review. This involves exposing potentially sensitive local content to a third-party endpoint for processing.
[PROMPT_INJECTION]: The skill processes untrusted content from external APIs (Twitter, web search, article extraction), creating a surface for indirect prompt injection where malicious instructions in the data could influence agent behavior.
Ingestion points: Responses from endpoints such as /api/twitter, /api/search, and /api/article on x402-api.onrender.com.
Boundary markers: No specific delimiters or instructions to ignore embedded commands are defined for the data returned by these tools.
Capability inventory: The agent has broad capabilities including wallet management, asset transfers, and executing shell commands via the CLI tool.
Sanitization: The instructions do not specify any validation or sanitization of the external content before it is processed by the agent.

Polygon Agent