The Agent Skills Directory

[Indirect Prompt Injection] (MEDIUM): The skill fetches drug-disease association data and 'content' from an external API (api.valyu.ai). This content is untrusted and could contain instructions designed to manipulate the agent's behavior when it processes the search results.
Ingestion points: Search results returned by scripts/search and scripts/search.mjs (specifically the content field).
Boundary markers: Absent. The documentation does not define delimiters to wrap the untrusted API content or instructions for the agent to ignore embedded commands.
Capability inventory: The skill executes local bash/Node.js scripts and performs network operations.
Sanitization: Absent. There is no evidence of filtering or escaping logic for the external data before it enters the agent context.
[Data Exposure & Exfiltration] (LOW): The skill performs network requests to api.valyu.ai using the Node.js fetch API. While this is the intended functionality, the domain is not on the trusted whitelist, creating a data channel to a third-party service.
[Command Execution] (LOW): The skill relies on executing a local wrapper script (scripts/search). The instructions suggest using a complex find command to resolve the script path in the user's home directory, which is a standard but notable local execution pattern.

open-targets-search