当请求需要真实 OpenDataWorks 平台能力时使用：元数据查询、表/字段发现、数据源路由、血缘、DDL、只读 SQL 验证/执行、结果格式化或图表契约输出。不用于业务语义或 NL2SQL 推理。
当请求需要 OpenDataWorks 平台通用业务语义时使用：元数据术语、工作流术语、血缘术语、平台指标定义、别名、歧义消解和业务规则例外。不用于领域专属本体、NL2SQL 方法或平台工具命令。
Use this built-in skill for chart construction and data storytelling from structured tabular results. The workflow is organized from the AntV chart-visualization playbook, but current default output is ECharts-compatible option JSON for local rendering.
Use this built-in skill for lightweight Python-based data exploration, profiling, aggregation, and file-based analysis. It does not fetch OpenDataWorks metadata on its own.
A minimal smoke skill used to verify that opendataagent can load bundled skills and surface them during chat.
Use this built-in skill for OpenDataWorks platform metadata, lineage, datasource resolution, and DDL lookup. This skill does not execute SQL for data analysis.
Use this built-in skill to execute readonly SQL through the OpenDataWorks agent API and shared odw-cli. Guardrails such as readonly checks, datasource scope, default limit, and timeout are enforced by CLI and API.
A minimal smoke skill used to verify that opendataagent can load bundled skills and surface them during chat.