The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions to download a standalone mining script (litcoin_miner.py) from the vendor's primary domain (litcoiin.xyz). It also specifies dependencies on the litcoin Python package and litcoin-mcp Node.js package.
[COMMAND_EXECUTION]: The documentation describes the process of running the downloaded Python script locally to initiate mining and reward claiming operations.
[REMOTE_CODE_EXECUTION]: The 'Research Mining' functionality implements a loop where the agent generates Python code to solve optimization problems and executes that code locally to verify performance against a baseline. This dynamic execution is the primary mechanism for 'Proof of Research'.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest external data (challenges and research tasks) from the api.litcoiin.xyz coordinator. This creates an attack surface for indirect prompt injection, where a compromised task could attempt to influence agent actions or access environment variables.
Ingestion points: Challenge and task descriptions are retrieved from the api.litcoiin.xyz endpoint.
Boundary markers: The documentation does not specify the use of delimiters or safety instructions when processing remote task data.
Capability inventory: The agent possesses the ability to execute code locally, perform network operations, and interact with the Base blockchain via the Bankr SDK.
Sanitization: No specific content sanitization or validation logic is defined for the remote tasks before they are processed by the agent.
[DATA_EXPOSURE]: The skill requires the BANKR_API_KEY environment variable for authentication and for managing DeFi operations such as staking, minting stablecoins, and claiming rewards.

litcoin-miner