The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected in the skill instructions or provided code snippets.
[COMMAND_EXECUTION]: The skill executes local scripts (scripts/search and scripts/search.mjs) to interface with the Valyu API.
It uses a dynamic path resolution method (find) to locate the execution script within the local plugin cache.
[EXTERNAL_DOWNLOADS]: The skill performs legitimate network operations to api.valyu.ai to fetch clinical trial data and refers users to platform.valyu.ai for API key management.
[CREDENTIALS_UNSAFE]: The skill includes a setup procedure for users to provide their Valyu API key via a CLI command (scripts/search setup <api-key>). This is a standard mechanism for API-reliant tools and does not involve hardcoded secrets.
[INDIRECT_PROMPT_INJECTION]: The skill processes clinical trial data from an external source (ClinicalTrials.gov via Valyu).
Ingestion points: Data is ingested through the results array in the JSON response from scripts/search (specifically the content field).
Boundary markers: The output is structured as JSON, which provides clear boundaries between data fields.
Capability inventory: The skill can execute local scripts and perform network requests to the designated API endpoint.
Sanitization: Results are returned in a structured JSON format to be parsed by the agent or tools like jq.

clinical-trials-search