The Agent Skills Directory

[COMMAND_EXECUTION]: The skill interacts with the local environment via shell commands to manage the trading lifecycle and communicate with the blockchain CLI tool.
Evidence: okx.py uses subprocess.run to call the onchainos CLI for operations such as wallet status checks, fetching market data, and broadcasting transactions.
Evidence: SKILL.md provides instructions for the agent to manage long-running background tasks using standard shell commands like nohup and ps.
[EXTERNAL_DOWNLOADS]: The system fetches market data from official vendor endpoints to drive its trading signals and backtesting engine.
Evidence: okx.py uses the standard Python urllib library to retrieve candle data from https://www.okx.com/api/v5/dex/market/candles.
[SAFE]: The skill incorporates robust security protocols for managing autonomous trading and high-risk operations.
Evidence: The 'Security: External Data Boundary' section in SKILL.md prevents the agent from interpreting API responses as instructions.
Evidence: The 'Live Trading Confirmation Protocol' ensures that real funds are never committed without the user's explicit session-level authorization.
Evidence: Transaction signing is performed in a Trusted Execution Environment (TEE) via the Agentic Wallet, ensuring private keys are never exposed to the local filesystem or logs.

mainstream-spot-order