The Agent Skills Directory

[PROMPT_INJECTION]: The skill ingests and processes untrusted market data (titles, descriptions, and tags) from the external DOME API, which creates a surface for indirect prompt injection.
Ingestion points: Market data enters the system through the fetchMarkets and searchMarkets functions in scripts/marketDiscovery.ts.
Boundary markers: The skill does not implement explicit boundary markers or instructions for the agent to ignore potentially malicious content within the fetched data.
Capability inventory: The skill possesses network-read capabilities via the fetch API. It does not have file-write, subprocess execution, or dynamic code execution capabilities.
Sanitization: No sanitization, escaping, or filtering of the external text content is performed before it is processed or returned.
[DATA_EXFILTRATION]: The skill performs network operations to a domain (api.domeapi.io) that is not on the default whitelist for data exfiltration analysis.
Evidence: In scripts/marketDiscovery.ts, the makeRequest function uses the global fetch API to communicate with https://api.domeapi.io/v1.

market-discovery