The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions for installing the '@makehq/endpoints-sdk' and '@makehq/sdk' packages from the npm registry. These are official packages maintained by the skill's author (Integromat) and are necessary for the skill's primary function.
[CREDENTIALS_UNSAFE]: The documentation contains strong security directives for secret management. It advises using project-local environment files (e.g., .env.local), explicitly warns against hardcoding API keys in source code or logs, and prohibits passing secrets via command-line arguments.
[INDIRECT_PROMPT_INJECTION]: The skill identifies data retrieved from external app actions as an ingestion point for untrusted content. It explicitly instructs the agent to treat provider responses as data rather than instructions and provides guidance on sanitizing terminal output to prevent the execution of malicious payloads.
[COMMAND_EXECUTION]: The skill utilizes CLI tools for discovering and executing app actions. It enforces security through a 'least privilege' approach, requiring explicit user approval for write operations and mandating strict schema validation before any command is executed.

make-endpoints