The Agent Skills Directory

[COMMAND_EXECUTION]: The strategy engine in rwa_alpha.py executes shell commands via subprocess.run to interact with the onchainos CLI for price discovery, wallet status checks, and trade execution. These interactions are standard for the skill's intended purpose.
[EXTERNAL_DOWNLOADS]: The skill fetches data from several well-known external services including Google News RSS for headlines, Polymarket API for prediction market probabilities, and the Anthropic API for event classification.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via the processing of untrusted external data.
Ingestion points: Financial headlines are ingested from Google News RSS feeds within rwa_alpha.py.
Boundary markers: The ingested headlines are processed by an LLM using a system prompt (_LLM_SYSTEM_PROMPT) designed to restrict output to specific event types.
Capability inventory: Classification results from the LLM trigger automated trading operations via the onchainos CLI, including market swaps and contract calls across Ethereum and Solana.
Sanitization: The skill employs keyword-based regex filtering before LLM classification and limits the model's output schema to a predefined list of macro events.

rwa-alpha