The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates downloading project templates from external Git repositories via the surrealkit init --from command. This behavior is documented as a core feature for scaffolding projects.
[COMMAND_EXECUTION]: The typegen and sync features support the execution of a user-defined shell command (such as a code formatter) specified in the surrealkit.toml configuration file.
[EXTERNAL_DOWNLOADS]: Installation instructions refer to official and well-known registries including GitHub Releases, Docker Hub (ghcr.io), and the Cargo package registry.
[PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection via the processing of remote Git templates and schema files.
Ingestion points: Remote Git repositories accessed via surrealkit init (documented in references/init-templates.md).
Boundary markers: No explicit delimiters are mentioned to separate untrusted template content from agent instructions.
Capability inventory: Includes file system writes, SQL execution (via sync, rollout, and seed), and HTTP requests through the api_request testing kind.
Sanitization: Relies on the underlying CLI tool's validation; no additional sanitization is defined in the skill instructions.
[SAFE]: The skill demonstrates secure credential handling by using environment variables for database and API authentication in the testing framework.

surrealkit