The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted blockchain data from an external gateway (gateway.usebuy.ai). This data, which includes transaction logs and instruction fields, could contain malicious natural language instructions designed to manipulate the agent's behavior during subsequent processing.
Ingestion points: External API endpoints defined in SKILL.md such as https://gateway.usebuy.ai/sqd/datasets/{dataset}/stream.
Boundary markers: The skill does not specify delimiters or instructions to ignore potential commands embedded within the retrieved blockchain data.
Capability inventory: The skill uses shell commands (curl) for data retrieval and exposes this data to the agent's context.
Sanitization: No sanitization or filtering logic is provided for the fetched data before it is parsed and processed.
[DATA_EXFILTRATION]: The skill performs network operations using curl and fetch to a domain (gateway.usebuy.ai) that is not on the standard whitelisted list of services. While currently used for retrieving public blockchain data, network access to non-whitelisted domains represents a possible communication channel for data transmission.

data-availability