The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill communicates with the external domain usecortx.dev to retrieve reliability metrics and monitoring status for x402 payment endpoints.- [COMMAND_EXECUTION]: The skill utilizes curl for API interaction and jq for parsing JSON responses. These are used strictly for the stated purpose of monitoring endpoint reliability.- [INDIRECT_PROMPT_INJECTION]: The skill processes data from a third-party API, which is a potential vector for indirect prompt injection. However, it incorporates strong defensive measures to mitigate this risk.
Ingestion points: JSON data is retrieved from usecortx.dev/api/v1/lookup and usecortx.dev/api/v1/reliability/ into temporary files (SKILL.md).
Boundary markers: The instructions explicitly require the agent to ignore any URLs, instructions, or payment requests found inside the response fields. A dedicated 'Security constraints' section reinforces that all string fields are to be treated as data only.
Capability inventory: The skill uses curl, mkdir, cat, and jq for data handling. It does not perform any execution or evaluation of the retrieved data.
Sanitization: Step 4 implements comprehensive validation, including data type checking, range validation for percentage fields, ISO 8601 date format verification, and strict matching of the endpoint_url against the user-provided input.

cortx-reliability