The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches the latest version string from the vendor's official domain at https://download.surrealdb.com.
[COMMAND_EXECUTION]: The skill uses surreal validate and npx @surrealdb/surql-fmt to provide query validation and formatting, which are official utilities from the vendor.
[CREDENTIALS_UNSAFE]: The references/schema.md file contains a documentation example that uses a placeholder password 'secret' in a DEFINE USER statement.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided queries through CLI tools, presenting a potential injection surface. * Ingestion points: SurrealQL queries provided by users for modification or troubleshooting in SKILL.md. * Boundary markers: The skill suggests using parameters to prevent database-level injection but does not define explicit shell-level delimiters for the external tools. * Capability inventory: The skill has the capability to execute shell commands like surreal and npx as part of its core functionality. * Sanitization: Reference documentation in references/querying.md promotes the use of parameterized queries as a best practice.

surrealql