The Agent Skills Directory

[COMMAND_EXECUTION]: The skill leverages a set of CLI tools (e.g., terse init, terse generate) specifically designed for managing Terse projects. These commands are used to automate project setup and deployment in a transparent manner.
[EXTERNAL_DOWNLOADS]: Installs the terse-cli utility via npm when necessary. As this is the official tool provided by the author ('terseai') for interacting with their platform, it is categorized as a safe vendor resource.
[CREDENTIALS_UNSAFE]: The skill provides detailed instructions for securely handling API keys and secrets via the platform's native secrets management system, encouraging best practices like inputting secrets via stdin and using base64 encoding for sensitive files.
[PROMPT_INJECTION]: To protect against potential injection in user-supplied workflow descriptions, the skill mandates the use of strict Zod schemas and agentic output validation during code generation.

terse-create