The Agent Skills Directory

[SAFE]: The skill facilitates read-only access to business data through the api_read tool. It incorporates robust instructional guardrails, such as requiring the agent to perform schema discovery via the meta method before querying and enforcing explicit row limits to prevent excessive data output in the chat environment.
[DYNAMIC_EXECUTION]: For data visualization, the skill generates self-contained SVG charts. This dynamic content generation is used exclusively for rendering charts within the chat and does not execute external code or make unauthorized network requests.
[SAFE]: The documentation explicitly forbids the use of data-modifying operations, such as api_write or api_delete, which are available on the underlying MCP server but are not utilized by this skill.

semantic-model