@clickzetta/cz-cli-darwin-arm64 0.3.20 → 0.3.21

package/bin/skills/clickzetta-access-control/eval_cases.jsonl

@@ -1,3 +1,3 @@
-{"case_id":"002","type":"should_call","user_input":"怎么查看当前工作空间有哪些用户？","expected_skill":"clickzetta-access-control","expected_output_contains":["SHOW USERS"]}
+{"case_id":"002","type":"should_call","user_input":"怎么管理工作空间的用户？新用户加入后需要授予什么角色？","expected_skill":"clickzetta-access-control","expected_output_contains":["角色","授予"]}
 {"case_id":"003","type":"should_call","user_input":"当前有哪些系统预置角色？各自有什么权限？","expected_skill":"clickzetta-access-control","expected_output_contains":["workspace_admin","workspace_dev"]}
 {"case_id":"004","type":"should_call","user_input":"怎么给某个用户授予 public schema 下所有表的只读权限？","expected_skill":"clickzetta-access-control","expected_output_contains":["GRANT","SELECT"]}

package/bin/skills/clickzetta-batch-sync-pipeline/eval_cases.jsonl

@@ -0,0 +1,5 @@
+{"case_id":"001","type":"should_call","user_input":"怎么创建离线同步任务把 MySQL 表定期同步到 Lakehouse？","expected_skill":"clickzetta-batch-sync-pipeline","expected_output_contains":["离线同步","Cron","调度"]}
+{"case_id":"002","type":"should_call","user_input":"多表离线同步支持自动建表吗？","expected_skill":"clickzetta-batch-sync-pipeline","expected_output_contains":["自动创建","多表"]}
+{"case_id":"003","type":"should_call","user_input":"离线同步的单表模式和多表模式怎么选？","expected_skill":"clickzetta-batch-sync-pipeline","expected_output_contains":["单表","多表","task_type"]}
+{"case_id":"004","type":"should_call","user_input":"批量同步支持 Schema Evolution 吗？新增字段会自动适配吗？","expected_skill":"clickzetta-batch-sync-pipeline","expected_output_contains":["Schema Evolution","多表"]}
+{"case_id":"005","type":"should_call","user_input":"怎么配置离线同步任务的调度周期？用 Cron 表达式吗？","expected_skill":"clickzetta-batch-sync-pipeline","expected_output_contains":["Cron","调度"]}

package/bin/skills/clickzetta-cdc-sync-pipeline/eval_cases.jsonl

@@ -0,0 +1,5 @@
+{"case_id":"001","type":"should_call","user_input":"怎么把 MySQL 整库实时同步到 Lakehouse？","expected_skill":"clickzetta-cdc-sync-pipeline","expected_output_contains":["整库镜像","Binlog","CDC"]}
+{"case_id":"002","type":"should_call","user_input":"分库分表的数据怎么合并同步到 Lakehouse 一张表？","expected_skill":"clickzetta-cdc-sync-pipeline","expected_output_contains":["多表合并","分库分表"]}
+{"case_id":"003","type":"should_call","user_input":"CDC 同步任务 Binlog 位点过期了怎么办？","expected_skill":"clickzetta-cdc-sync-pipeline","expected_output_contains":["Binlog"]}
+{"case_id":"004","type":"should_call","user_input":"多表实时同步怎么配置告警？支持飞书通知吗？","expected_skill":"clickzetta-cdc-sync-pipeline","expected_output_contains":["告警","webhook"]}
+{"case_id":"005","type":"should_call","user_input":"PostgreSQL 整库 CDC 同步需要源端做什么准备？","expected_skill":"clickzetta-cdc-sync-pipeline","expected_output_contains":["WAL","PostgreSQL"]}

package/bin/skills/clickzetta-dba-guide/SKILL.md

@@ -0,0 +1,542 @@
+---
+name: clickzetta-dba-guide
+description: |
+  ClickZetta Lakehouse DBA 日常运维操作手册。集中覆盖 DBA 最常用的 6 类操作：
+  计算集群运维、作业监控与诊断、数据恢复与保护、
+  存储优化与维护、Schema 与对象管理、成本与资源分析。
+  每个操作提供可直接执行的 SQL，并标注 ClickZetta 特有限制。
+  当用户说"启停集群"、"调整集群规格"、
+  "取消作业"、"慢查询"、"恢复误删表"、"UNDROP"、"RESTORE"、
+  "小文件合并"、"OPTIMIZE"、"ANALYZE TABLE"、
+  "成本分析"、"存储用量"、"DBA 操作"、
+  "创建 Schema"、"删除 Schema"、"重命名表"、"对象管理"、"Schema 管理"时触发。
+  用户与权限管理、网络策略、数据脱敏等安全治理操作请使用 clickzetta-access-control skill。
+  Keywords: DBA, operations, monitoring, troubleshooting, cluster management, cost
+---
+
+# ClickZetta Lakehouse DBA 运维手册
+
+---
+
+## 模块 1：用户与权限管理
+
+### 用户管理
+
+```sql
+-- 创建用户（设置默认集群和 Schema）
+CREATE USER alice DEFAULT_VCLUSTER default_ap DEFAULT_SCHEMA my_schema;
+
+-- 修改用户默认集群
+ALTER USER alice SET DEFAULT_VCLUSTER = analytics_cluster;
+
+-- 删除用户（从当前工作空间移除）
+DROP USER alice;
+
+-- 查看所有用户
+SHOW USERS;
+```
+
+### 角色管理
+
+```sql
+-- 创建自定义角色（仅工作空间级，仅 SQL）
+CREATE ROLE data_engineer;
+
+-- 将角色授予用户
+GRANT ROLE data_engineer TO USER alice;
+
+-- 撤销角色
+REVOKE ROLE data_engineer FROM USER alice;
+
+-- 查看所有角色
+SHOW ROLES;
+
+-- 查看用户权限
+SHOW GRANTS TO USER alice;
+SHOW GRANTS TO ROLE data_engineer;
+```
+
+### 权限授予
+
+```sql
+-- 授予 Schema 下所有表的读权限
+GRANT SELECT ON ALL TABLES IN SCHEMA my_schema TO ROLE data_engineer;
+
+-- 授予单张表的读写权限
+GRANT SELECT, INSERT, UPDATE, DELETE ON TABLE my_schema.orders TO USER alice;
+
+-- 授予创建表的权限
+GRANT CREATE TABLE ON SCHEMA my_schema TO ROLE data_engineer;
+
+-- 授予使用集群的权限
+GRANT USE ON VCLUSTER default_ap TO ROLE data_engineer;
+
+-- 授予 information_schema 查询权限
+GRANT ALL ON ALL VIEWS IN SCHEMA information_schema TO ROLE data_engineer;
+
+-- 批量授权（Schema 级别）
+GRANT SELECT ON ALL TABLES IN SCHEMA ods TO ROLE analyst;
+GRANT SELECT ON ALL TABLES IN SCHEMA dwd TO ROLE analyst;
+GRANT SELECT ON ALL TABLES IN SCHEMA dws TO ROLE analyst;
+```
+
+### 权限撤销
+
+```sql
+-- 撤销表权限
+REVOKE SELECT ON TABLE my_schema.orders FROM USER alice;
+
+-- 撤销 Schema 创建权限
+REVOKE CREATE TABLE ON SCHEMA my_schema FROM ROLE data_engineer;
+```
+
+### 动态脱敏（列级安全，邀测功能）
+
+```sql
+-- 创建脱敏函数（基于角色）
+CREATE FUNCTION my_schema.phone_masking(phone STRING)
+RETURNS STRING
+AS CASE
+  WHEN ARRAY_CONTAINS(current_roles(), 'data_admin') THEN phone
+  ELSE CONCAT(SUBSTR(phone, 1, 3), '****', SUBSTR(phone, 8, 4))
+END;
+
+-- 绑定脱敏策略到列
+ALTER TABLE my_schema.users
+CHANGE COLUMN phone SET MASK my_schema.phone_masking;
+
+-- 解除脱敏
+ALTER TABLE my_schema.users
+CHANGE COLUMN phone UNSET MASK;
+```
+
+**ClickZetta 特有限制：**
+- 无超级用户，所有操作必须明确授权
+- `instance_admin` 不能直接操作工作空间数据
+- 自定义角色仅工作空间级，不支持实例级自定义角色
+
+---
+
+## 模块 2：计算集群运维
+
+### 启停与状态
+
+```sql
+-- 启动集群
+ALTER VCLUSTER my_cluster RESUME;
+ALTER VCLUSTER IF EXISTS my_cluster RESUME;
+
+-- 停止集群
+ALTER VCLUSTER my_cluster SUSPEND;
+ALTER VCLUSTER my_cluster SUSPEND FORCE;  -- 强制停止（中断运行中的作业）
+
+-- 取消集群所有作业
+ALTER VCLUSTER my_cluster CANCEL ALL JOBS;
+
+-- 查看集群状态
+SHOW VCLUSTERS;
+SHOW VCLUSTERS WHERE state = 'RUNNING';
+SHOW VCLUSTERS WHERE state = 'SUSPENDED';
+DESC VCLUSTER my_cluster;
+DESC VCLUSTER EXTENDED my_cluster;
+
+-- 切换当前会话使用的集群
+USE VCLUSTER my_cluster;
+```
+
+### 调整规格
+
+```sql
+-- 通用型（GP）：固定规格
+ALTER VCLUSTER my_gp SET VCLUSTER_SIZE = 8;
+
+-- 通用型（GP）：弹性规格
+ALTER VCLUSTER my_gp SET MIN_VCLUSTER_SIZE = 2 MAX_VCLUSTER_SIZE = 16;
+
+-- 分析型（AP）：调整副本数
+ALTER VCLUSTER my_ap SET MIN_REPLICAS = 1 MAX_REPLICAS = 4;
+
+-- 分析型（AP）：调整最大并发
+ALTER VCLUSTER my_ap SET MAX_CONCURRENCY = 16;
+
+-- 设置查询超时（秒，-1 表示无限制）
+ALTER VCLUSTER my_cluster SET QUERY_RUNTIME_LIMIT_IN_SECOND = 3600;
+```
+
+### 自动停止与启动
+
+```sql
+-- 设置 60 秒无作业自动停止，有作业自动启动
+ALTER VCLUSTER my_cluster SET
+  AUTO_SUSPEND_IN_SECOND = 60
+  AUTO_RESUME = TRUE;
+
+-- 关闭自动停止
+ALTER VCLUSTER my_cluster SET AUTO_SUSPEND_IN_SECOND = -1;
+```
+
+### AP 集群预加载缓存
+
+```sql
+-- 设置预加载表（集群启动时自动缓存最新数据）
+ALTER VCLUSTER my_ap SET PRELOAD_TABLES = "sales.orders,sales.products";
+
+-- 查看缓存状态
+SHOW PRELOAD CACHED STATUS;
+SHOW EXTENDED PRELOAD CACHED STATUS;
+```
+
+**ClickZetta 特有限制：**
+- OPTIMIZE（小文件合并）仅 GP 集群支持，AP 集群不生效
+- 分析型集群规格步长为 2^n（1/2/4/8/16...），通用型步长为 1
+
+---
+
+## 模块 3：作业监控与诊断
+
+### 实时作业查看
+
+```sql
+-- 查看最近作业（最多 7 天，10000 条）
+SHOW JOBS LIMIT 20;
+SHOW JOBS IN VCLUSTER default_ap LIMIT 20;
+
+-- 取消指定作业
+CANCEL JOB '2026050118342658136171272';
+
+-- 查看执行计划
+EXPLAIN SELECT * FROM orders WHERE order_date = '2024-01-01';
+EXPLAIN EXTENDED SELECT * FROM orders WHERE order_date = '2024-01-01';
+```
+
+### 历史作业分析（information_schema）
+
+```sql
+-- 慢查询 TOP 20（最近 7 天）
+SELECT job_id, job_creator, execution_time, input_bytes, job_text
+FROM information_schema.job_history
+WHERE pt_date >= CAST(CURRENT_DATE - INTERVAL 7 DAY AS DATE)
+  AND status = 'SUCCEED'
+ORDER BY execution_time DESC
+LIMIT 20;
+
+-- 失败作业（最近 24 小时）
+SELECT job_id, job_creator, error_message, start_time, job_text
+FROM information_schema.job_history
+WHERE pt_date >= CAST(CURRENT_DATE - INTERVAL 1 DAY AS DATE)
+  AND status = 'FAILED'
+ORDER BY start_time DESC;
+
+-- 按用户统计 CRU 消耗（最近 30 天）
+SELECT job_creator,
+       COUNT(*) AS job_count,
+       ROUND(SUM(cru), 2) AS total_cru,
+       ROUND(AVG(execution_time), 1) AS avg_exec_sec
+FROM information_schema.job_history
+WHERE pt_date >= CAST(CURRENT_DATE - INTERVAL 30 DAY AS DATE)
+  AND status = 'SUCCEED'
+GROUP BY job_creator
+ORDER BY total_cru DESC;
+
+-- 按集群统计作业分布
+SELECT virtual_cluster,
+       COUNT(*) AS job_count,
+       ROUND(SUM(cru), 2) AS total_cru
+FROM information_schema.job_history
+WHERE pt_date >= CAST(CURRENT_DATE - INTERVAL 7 DAY AS DATE)
+GROUP BY virtual_cluster
+ORDER BY total_cru DESC;
+```
+
+---
+
+## 模块 4：数据恢复与保护
+
+### 恢复误删对象
+
+```sql
+-- 查看已删除的表（delete_time 不为 NULL）
+SHOW TABLES HISTORY IN my_schema;
+SHOW TABLES HISTORY LIKE '%orders%';
+
+-- 恢复误删的表/动态表/物化视图
+UNDROP TABLE my_schema.orders;
+UNDROP TABLE my_schema.my_dynamic_table;
+UNDROP TABLE my_schema.my_mv;
+-- ⚠️ 恢复外部函数用 UNDROP FUNCTION，不是 UNDROP EXTERNAL FUNCTION
+UNDROP FUNCTION my_schema.my_ext_function;
+```
+
+### 回滚到历史版本
+
+```sql
+-- 查看表的版本历史
+DESC HISTORY my_schema.orders;
+-- 返回：version, time, total_rows, total_bytes, user, operation, job_id
+
+-- 恢复到指定时间点（覆盖当前数据）
+-- ⚠️ 时间戳必须用 CAST() 或完整毫秒格式，不能用简单字符串
+-- ❌ 错误：RESTORE TABLE t TO TIMESTAMP AS OF '2024-01-15';
+-- ✅ 正确写法：
+RESTORE TABLE my_schema.orders TO TIMESTAMP AS OF CAST('2024-01-15 10:00:00' AS TIMESTAMP);
+RESTORE TABLE my_schema.orders TO TIMESTAMP AS OF CURRENT_TIMESTAMP() - INTERVAL '2' HOURS;
+-- 也支持完整毫秒时间戳字符串（从 DESC HISTORY 复制）：
+RESTORE TABLE my_schema.orders TO TIMESTAMP AS OF '2024-01-15 10:00:00.123';
+
+-- 查询历史数据（不覆盖，仅查看）
+SELECT * FROM my_schema.orders TIMESTAMP AS OF CAST('2024-01-15 10:00:00' AS TIMESTAMP);
+```
+
+### 设置数据保留周期
+
+```sql
+-- 设置 Time Travel 保留 30 天（范围 0-90）
+ALTER TABLE my_schema.orders SET PROPERTIES ('data_retention_days' = '30');
+
+-- 查看当前设置
+SHOW CREATE TABLE my_schema.orders;
+```
+
+**ClickZetta 特有限制：**
+- `RESTORE TABLE` 目标时间点不能早于表创建时间
+- `UNDROP` 需在 `data_retention_days` 保留期内（默认 1 天）
+- 物化视图支持 UNDROP，但不支持 RESTORE
+
+---
+
+## 模块 5：存储优化与维护
+
+### 小文件合并
+
+```sql
+-- 手动触发小文件合并（异步，仅 GP 集群）
+OPTIMIZE my_schema.orders;
+
+-- 同步执行（等待完成）
+OPTIMIZE my_schema.orders OPTIONS('cz.sql.optimize.table.async' = 'false');
+
+-- 只优化特定分区
+OPTIMIZE my_schema.orders WHERE dt = '2024-01-01';
+OPTIMIZE my_schema.orders WHERE dt = '2024-01-01' AND region = 'cn';
+
+-- DML 写入时自动触发合并（GP 集群）
+SET cz.sql.compaction.after.commit = true;
+INSERT INTO my_schema.orders SELECT * FROM staging;
+```
+
+### 统计信息收集
+
+```sql
+-- 收集表统计信息（优化查询计划）
+ANALYZE TABLE my_schema.orders COMPUTE STATISTICS;
+
+-- 仅收集大小，不扫描数据（快速）
+ANALYZE TABLE my_schema.orders COMPUTE STATISTICS NOSCAN;
+
+-- 收集指定列的统计信息
+ANALYZE TABLE my_schema.orders COMPUTE STATISTICS FOR COLUMNS order_date, customer_id;
+
+-- 收集 Schema 下所有表
+ANALYZE TABLES IN my_schema COMPUTE STATISTICS;
+```
+
+### 清空数据
+
+```sql
+-- 清空整张表（保留表结构）
+TRUNCATE TABLE my_schema.staging;
+
+-- 清空指定分区
+TRUNCATE TABLE my_schema.orders WHERE dt = '2024-01-01';
+```
+
+### 查看存储用量
+
+```sql
+-- 当前 Schema 下大表排行
+SELECT table_schema, table_name,
+       ROUND(bytes / 1024.0 / 1024 / 1024, 2) AS size_gb,
+       row_count
+FROM information_schema.tables
+WHERE table_type = 'MANAGED_TABLE'
+ORDER BY bytes DESC
+LIMIT 20;
+
+-- Sort Key 推荐（系统自动分析）
+SELECT table_name, col, statement, ratio
+FROM information_schema.sortkey_candidates
+ORDER BY ratio DESC;
+```
+
+---
+
+## 模块 6：网络策略与安全
+
+### 网络策略管理
+
+```sql
+-- 创建网络策略（白名单）
+CREATE NETWORK POLICY office_policy
+  ALLOWED_IP_LIST = ('10.0.0.0/8', '192.168.1.0/24')
+  COMMENT '办公网络白名单';
+
+-- 创建网络策略（白名单 + 黑名单）
+CREATE NETWORK POLICY strict_policy
+  ALLOWED_IP_LIST = ('10.0.0.0/8')
+  BLOCKED_IP_LIST = ('10.0.1.100')
+  COMMENT '严格访问控制';
+
+-- 修改网络策略（覆盖式，必须包含所有 IP）
+ALTER NETWORK POLICY office_policy
+  ALLOWED_IP_LIST = ('10.0.0.0/8', '172.16.0.0/12')
+  BLOCKED_IP_LIST = ('10.0.1.100');
+
+-- 停用/启用策略
+ALTER NETWORK POLICY office_policy INACTIVATE;
+ALTER NETWORK POLICY office_policy ACTIVATE;
+
+-- 删除策略
+DROP NETWORK POLICY IF EXISTS office_policy;
+
+-- 查看所有策略（注意：单数 POLICY，无 S）
+SHOW NETWORK POLICY;
+
+-- 查看策略详情
+DESC NETWORK POLICY office_policy;
+```
+
+**关键规则（Deny 优先）：**
+- 无任何策略时：允许所有 IP
+- 有白名单策略时：不在白名单的 IP 被拒绝
+- 黑名单命中时：无论白名单如何，该 IP 被拒绝
+- MySQL 协议：只要有任何生效策略，所有 MySQL 流量均被拦截
+- 策略生效延迟：最多 5 分钟
+
+---
+
+## 模块 7：Schema 与对象管理
+
+### Schema 管理
+
+```sql
+-- 创建 Schema
+CREATE SCHEMA ods;
+CREATE SCHEMA IF NOT EXISTS dwd;
+
+-- 修改 Schema 注释
+ALTER SCHEMA ods SET COMMENT 'ODS 原始数据层';
+
+-- 重命名 Schema
+ALTER SCHEMA old_name RENAME TO new_name;
+
+-- 删除 Schema（级联删除所有对象）
+DROP SCHEMA IF EXISTS temp_schema CASCADE;
+
+-- 切换默认 Schema
+USE SCHEMA my_schema;
+```
+
+### 表管理
+
+```sql
+-- 修改表：加列
+ALTER TABLE my_schema.orders ADD COLUMN (discount DECIMAL(5,2) COMMENT '折扣率');
+
+-- 修改表：改列注释
+ALTER TABLE my_schema.orders CHANGE COLUMN order_id SET COMMENT '订单唯一标识';
+
+-- 修改表：设置生命周期
+ALTER TABLE my_schema.orders SET PROPERTIES ('data_lifecycle' = '90');
+
+-- 修改表：设置 Sort Key
+ALTER TABLE my_schema.orders SET PROPERTIES ('hint.sort.columns' = 'order_date');
+
+-- 重命名表
+ALTER TABLE my_schema.orders RENAME TO my_schema.orders_v2;
+
+-- 删除表（可 UNDROP 恢复）
+DROP TABLE IF EXISTS my_schema.temp_table;
+
+-- 删除动态表
+DROP DYNAMIC TABLE IF EXISTS my_schema.my_dt;
+
+-- 删除物化视图
+DROP MATERIALIZED VIEW IF EXISTS my_schema.my_mv;
+```
+
+### 批量对象查看
+
+```sql
+-- 统计各类型对象数量
+SELECT
+  CASE WHEN is_view THEN 'VIEW'
+       WHEN is_materialized_view THEN 'MV'
+       WHEN is_dynamic THEN 'DT'
+       WHEN is_external THEN 'EXTERNAL'
+       ELSE 'TABLE' END AS type,
+  COUNT(*) AS cnt
+FROM (SHOW TABLES IN my_schema)
+GROUP BY 1;
+
+-- 查找大于 30 天未更新的表（潜在废弃表）
+SELECT table_schema, table_name, last_modify_time,
+       ROUND(bytes / 1024.0 / 1024 / 1024, 2) AS size_gb
+FROM information_schema.tables
+WHERE table_type = 'MANAGED_TABLE'
+  AND last_modify_time < CURRENT_TIMESTAMP - INTERVAL 30 DAY
+ORDER BY bytes DESC;
+```
+
+---
+
+## 模块 8：成本与资源分析（需 INSTANCE ADMIN）
+
+```sql
+-- 本月各工作空间计算费用
+SELECT workspace_name, sku_name,
+       ROUND(SUM(measurements_consumption), 2) AS total_cru,
+       ROUND(SUM(amount), 2) AS total_yuan
+FROM SYS.information_schema.instance_usage
+WHERE measurement_start >= DATE_TRUNC('month', CURRENT_DATE)
+GROUP BY workspace_name, sku_name
+ORDER BY total_yuan DESC;
+
+-- 本月各工作空间存储费用
+SELECT workspace_name, sku_name,
+       ROUND(SUM(measurements_consumption), 4) AS consumption,
+       measurements_unit,
+       ROUND(SUM(amount), 4) AS total_yuan
+FROM SYS.information_schema.storage_metering
+WHERE measurement_start >= DATE_TRUNC('month', CURRENT_DATE)
+GROUP BY workspace_name, sku_name, measurements_unit
+ORDER BY workspace_name, total_yuan DESC;
+
+-- 跨空间存储用量排行
+SELECT workspace_name,
+       ROUND(workspace_storage / 1024.0 / 1024 / 1024, 2) AS storage_gb
+FROM SYS.information_schema.workspaces
+WHERE delete_time IS NULL
+ORDER BY workspace_storage DESC;
+
+-- 跨空间大表排行（大于 10GB）
+SELECT table_catalog, table_schema, table_name,
+       ROUND(bytes / 1024.0 / 1024 / 1024, 2) AS size_gb, row_count
+FROM SYS.information_schema.tables
+WHERE delete_time IS NULL AND bytes > 10 * 1024 * 1024 * 1024
+ORDER BY bytes DESC;
+```
+
+---
+
+## ClickZetta DBA 特有注意事项
+
+| 场景 | 注意事项 |
+|---|---|
+| 权限体系 | 无超级用户；`instance_admin` 不能直接操作工作空间数据 |
+| 自定义角色 | 仅工作空间级，不支持实例级；只能 SQL 创建，不支持 Web 端 |
+| OPTIMIZE | 仅 GP 集群支持；AP 集群不支持小文件合并 |
+| UNDROP | 需在 `data_retention_days` 保留期内（默认 1 天） |
+| RESTORE | 目标时间点不能早于表创建时间 |
+| 网络策略 | Deny 优先；MySQL 协议有任何策略即全部拦截；生效延迟最多 5 分钟 |
+| 动态脱敏 | 邀测功能，需联系技术支持开通 |
+| 集群规格 | AP 集群步长 2^n；GP 集群步长 1；同步型最小 0.25 CRU |

package/bin/skills/clickzetta-dba-guide/eval_cases.jsonl

@@ -0,0 +1,3 @@
+{"case_id":"001","type":"should_call","user_input":"DBA 日常怎么启停集群和调整规格？","expected_skill":"clickzetta-dba-guide","expected_output_contains":["SUSPEND","RESUME"]}
+{"case_id":"002","type":"should_call","user_input":"怎么取消正在运行的作业？CANCEL JOB 怎么用？","expected_skill":"clickzetta-dba-guide","expected_output_contains":["CANCEL"]}
+{"case_id":"003","type":"should_call","user_input":"DBA 日常怎么做存储优化？小文件合并和 OPTIMIZE 怎么用？","expected_skill":"clickzetta-dba-guide","expected_output_contains":["OPTIMIZE","小文件"]}

package/bin/skills/clickzetta-dw-modeling/eval_cases.jsonl

@@ -1,4 +1,4 @@
-{"case_id":"001","type":"should_call","user_input":"帮我做数仓分层设计","expected_skill":"clickzetta-dw-modeling","expected_output_contains":["分层"]}
+{"case_id":"001","type":"should_call","user_input":"数仓分层设计的原则是什么？ODS、DWD、DWS 各层的职责？","expected_skill":"clickzetta-dw-modeling","expected_output_contains":["分层"]}
 {"case_id":"002","type":"should_call","user_input":"ODS/DWD/DWS/ADS 分层怎么设计","expected_skill":"clickzetta-dw-modeling","expected_output_contains":["ODS","DWD","DWS"]}
 {"case_id":"003","type":"should_call","user_input":"Medallion 架构 Bronze/Silver/Gold 怎么搭建","expected_skill":"clickzetta-dw-modeling","expected_output_contains":["Bronze","Silver","Gold"]}
 {"case_id":"004","type":"should_call","user_input":"星型模型和雪花模型怎么选","expected_skill":"clickzetta-dw-modeling","expected_output_contains":["星型","雪花"]}

package/bin/skills/clickzetta-dynamic-table/eval_cases.jsonl

@@ -0,0 +1,5 @@
+{"case_id":"001","type":"should_call","user_input":"帮我创建一个 Dynamic Table，从 public.dim_studio_user_dmin_f 聚合按租户+日期统计用户数，每 60 分钟自动刷新","expected_skill":"clickzetta-dynamic-table","expected_output_contains":["DYNAMIC TABLE","REFRESH"]}
+{"case_id":"002","type":"should_call","user_input":"怎么查看动态表的刷新历史和状态","expected_skill":"clickzetta-dynamic-table","expected_output_contains":["REFRESH HISTORY"]}
+{"case_id":"003","type":"should_call","user_input":"动态表的增量刷新怎么配置？SESSION_CONFIGS 怎么用？","expected_skill":"clickzetta-dynamic-table","expected_output_contains":["SESSION_CONFIGS","增量"]}
+{"case_id":"004","type":"should_call","user_input":"静态分区 DT 和动态分区 DT 有什么区别？该怎么选？","expected_skill":"clickzetta-dynamic-table","expected_output_contains":["静态分区","动态分区"]}
+{"case_id":"005","type":"should_call","user_input":"动态表怎么修改刷新间隔和 vcluster？","expected_skill":"clickzetta-dynamic-table","expected_output_contains":["ALTER","DYNAMIC TABLE"]}

package/bin/skills/clickzetta-file-import-pipeline/eval_cases.jsonl

@@ -0,0 +1,5 @@
+{"case_id":"001","type":"should_call","user_input":"从 URL 导入文件到 Lakehouse 的步骤和语法是什么？","expected_skill":"clickzetta-file-import-pipeline","expected_output_contains":["Volume","COPY INTO"]}
+{"case_id":"002","type":"should_call","user_input":"本地文件上传到 Lakehouse 表的流程是什么？需要哪些步骤？","expected_skill":"clickzetta-file-import-pipeline","expected_output_contains":["Volume","COPY INTO"]}
+{"case_id":"003","type":"should_call","user_input":"COPY INTO 导入数据时 append 和 overwrite 写入模式有什么区别？请说明语法","expected_skill":"clickzetta-file-import-pipeline","expected_output_contains":["append","overwrite"]}
+{"case_id":"004","type":"should_call","user_input":"COPY INTO 导入前怎么推断文件格式？有哪些支持的格式类型？","expected_skill":"clickzetta-file-import-pipeline","expected_output_contains":["CSV","JSON","Parquet"]}
+{"case_id":"005","type":"should_call","user_input":"CSV 有自定义分隔符，COPY INTO 的 OPTIONS 怎么写？","expected_skill":"clickzetta-file-import-pipeline","expected_output_contains":["CSV","OPTIONS"]}

package/bin/skills/clickzetta-oss-ingest-pipeline/eval_cases.jsonl

@@ -0,0 +1,5 @@
+{"case_id":"001","type":"should_call","user_input":"怎么从阿里云 OSS 持续自动导入数据到 Lakehouse？","expected_skill":"clickzetta-oss-ingest-pipeline","expected_output_contains":["PIPE","LIST_PURGE"]}
+{"case_id":"002","type":"should_call","user_input":"OSS PIPE 的 LIST_PURGE 和 EVENT_NOTIFICATION 模式有什么区别？","expected_skill":"clickzetta-oss-ingest-pipeline","expected_output_contains":["LIST_PURGE","EVENT_NOTIFICATION"]}
+{"case_id":"003","type":"should_call","user_input":"怎么从 S3 批量导入 Parquet 文件到 Lakehouse？","expected_skill":"clickzetta-oss-ingest-pipeline","expected_output_contains":["Volume","COPY INTO"]}
+{"case_id":"004","type":"should_call","user_input":"OSS 持续导入的前置步骤是什么？需要先创建什么对象？","expected_skill":"clickzetta-oss-ingest-pipeline","expected_output_contains":["CREATE STORAGE CONNECTION","External Volume"]}
+{"case_id":"005","type":"should_call","user_input":"腾讯云 COS 的数据怎么导入 ClickZetta？","expected_skill":"clickzetta-oss-ingest-pipeline","expected_output_contains":["COS","PIPE"]}

package/bin/skills/clickzetta-realtime-sync-pipeline/eval_cases.jsonl

@@ -0,0 +1,5 @@
+{"case_id":"001","type":"should_call","user_input":"怎么用 Studio 创建单表实时同步任务？","expected_skill":"clickzetta-realtime-sync-pipeline","expected_output_contains":["实时同步","task_type","28"]}
+{"case_id":"002","type":"should_call","user_input":"Kafka 单个 topic 实时同步到 Lakehouse 表怎么配置？","expected_skill":"clickzetta-realtime-sync-pipeline","expected_output_contains":["Kafka","实时同步"]}
+{"case_id":"003","type":"should_call","user_input":"单表实时同步和多表实时同步有什么区别？","expected_skill":"clickzetta-realtime-sync-pipeline","expected_output_contains":["单表","多表","28","281"]}
+{"case_id":"004","type":"should_call","user_input":"MySQL 单表 CDC 实时同步到 Lakehouse 怎么做？","expected_skill":"clickzetta-realtime-sync-pipeline","expected_output_contains":["MySQL","实时同步","CDC"]}
+{"case_id":"005","type":"should_call","user_input":"实时同步任务需要配置调度策略吗？","expected_skill":"clickzetta-realtime-sync-pipeline","expected_output_contains":["无需配置","持续运行"]}

package/bin/skills/clickzetta-sql-pipeline-manager/eval_cases.jsonl

@@ -0,0 +1,12 @@
+{"case_id":"001","type":"should_call","user_input":"帮我创建一个动态表，每 5 分钟从 raw_events 聚合数据","expected_skill":"clickzetta-sql-pipeline-manager","expected_output_contains":["CREATE DYNAMIC TABLE","REFRESH INTERVAL"]}
+{"case_id":"002","type":"should_call","user_input":"怎么创建物化视图？","expected_skill":"clickzetta-sql-pipeline-manager","expected_output_contains":["CREATE MATERIALIZED VIEW"]}
+{"case_id":"003","type":"should_call","user_input":"创建一个 Table Stream 捕获 orders 表的变更","expected_skill":"clickzetta-sql-pipeline-manager","expected_output_contains":["CREATE TABLE STREAM"]}
+{"case_id":"004","type":"should_call","user_input":"怎么暂停动态表的刷新","expected_skill":"clickzetta-sql-pipeline-manager","expected_output_contains":["ALTER","SUSPEND"]}
+{"case_id":"005","type":"should_call","user_input":"怎么查看动态表的刷新历史","expected_skill":"clickzetta-sql-pipeline-manager","expected_output_contains":["SHOW DYNAMIC TABLE REFRESH HISTORY"]}
+{"case_id":"006","type":"should_call","user_input":"帮我设计一个 Medallion 架构的数据管道","expected_skill":"clickzetta-sql-pipeline-manager","expected_output_contains":["Bronze","Silver","Gold"]}
+{"case_id":"007","type":"should_call","user_input":"从 Kafka 持续导入数据到 Lakehouse 用什么方式","expected_skill":"clickzetta-sql-pipeline-manager","expected_output_contains":["Pipe","read_kafka"]}
+{"case_id":"008","type":"should_not_call","user_input":"帮我写一个 Node.js 后端","forbidden_skill":"clickzetta-sql-pipeline-manager"}
+{"case_id":"009","type":"should_not_call","user_input":"怎么创建用户和授权","forbidden_skill":"clickzetta-sql-pipeline-manager"}
+{"case_id":"010","type":"should_not_call","user_input":"Kubernetes 怎么部署","forbidden_skill":"clickzetta-sql-pipeline-manager"}
+{"case_id":"011","type":"should_not_call","user_input":"怎么连接 Superset","forbidden_skill":"clickzetta-sql-pipeline-manager"}
+{"case_id":"012","type":"should_not_call","user_input":"帮我优化一个慢查询","forbidden_skill":"clickzetta-sql-pipeline-manager"}

package/bin/skills/clickzetta-table-stream-pipeline/eval_cases.jsonl

@@ -0,0 +1,5 @@
+{"case_id":"001","type":"should_call","user_input":"怎么创建 Table Stream 捕获表的增量变更？","expected_skill":"clickzetta-table-stream-pipeline","expected_output_contains":["CREATE TABLE STREAM","change_tracking"]}
+{"case_id":"002","type":"should_call","user_input":"Table Stream 的 STANDARD 和 APPEND_ONLY 模式有什么区别？","expected_skill":"clickzetta-table-stream-pipeline","expected_output_contains":["STANDARD","APPEND_ONLY","INSERT"]}
+{"case_id":"003","type":"should_call","user_input":"Table Stream 消费后 offset 怎么管理？","expected_skill":"clickzetta-table-stream-pipeline","expected_output_contains":["offset"]}
+{"case_id":"004","type":"should_call","user_input":"怎么用 Table Stream + MERGE 实现幂等增量消费？","expected_skill":"clickzetta-table-stream-pipeline","expected_output_contains":["MERGE","幂等"]}
+{"case_id":"005","type":"should_call","user_input":"Table Stream 的 __change_type 元数据字段有哪些值？","expected_skill":"clickzetta-table-stream-pipeline","expected_output_contains":["__change_type","INSERT","UPDATE","DELETE"]}

package/bin/skills/clickzetta-vcluster-manager/eval_cases.jsonl

@@ -0,0 +1,5 @@
+{"case_id":"001","type":"should_call","user_input":"创建分析型集群的语法是什么？需要哪些参数？","expected_skill":"clickzetta-vcluster-manager","expected_output_contains":["CREATE VCLUSTER","ANALYTICS","VCLUSTER_SIZE"]}
+{"case_id":"002","type":"should_call","user_input":"VCluster 的三种集群类型分别适合什么场景？","expected_skill":"clickzetta-vcluster-manager","expected_output_contains":["GENERAL","ANALYTICS","INTEGRATION"]}
+{"case_id":"003","type":"should_call","user_input":"分析型集群的副本扩缩容参数怎么配置？","expected_skill":"clickzetta-vcluster-manager","expected_output_contains":["MIN_REPLICAS","MAX_REPLICAS"]}
+{"case_id":"004","type":"should_call","user_input":"PRELOAD_TABLES 缓存预加载的语法和限制是什么？","expected_skill":"clickzetta-vcluster-manager","expected_output_contains":["PRELOAD_TABLES","ALTER VCLUSTER"]}
+{"case_id":"005","type":"should_call","user_input":"集群的 AUTO_SUSPEND 和 AUTO_RESUME 机制是怎样的？","expected_skill":"clickzetta-vcluster-manager","expected_output_contains":["AUTO_SUSPEND_IN_SECOND","AUTO_RESUME"]}

package/bin/skills/clickzetta-volume-manager/eval_cases.jsonl

@@ -0,0 +1,5 @@
+{"case_id":"001","type":"should_call","user_input":"我要挂载一个阿里云 OSS bucket 到 Lakehouse，需要先创建 Storage Connection，endpoint 是 oss-cn-hangzhou-internal.aliyuncs.com，帮我说明完整流程","expected_skill":"clickzetta-volume-manager","expected_output_contains":["connection","volume"]}
+{"case_id":"002","type":"should_call","user_input":"创建一个外部 Volume 叫 eval_oss_volume，挂载 oss://studio-dev-hz/，用 eval_oss_conn 连接，开启目录自动刷新","expected_skill":"clickzetta-volume-manager","expected_output_contains":["eval_oss_volume"]}
+{"case_id":"003","type":"should_call","user_input":"查看 eval_oss_volume 里有哪些文件","expected_skill":"clickzetta-volume-manager","expected_output_contains":["eval_oss_volume"]}
+{"case_id":"004","type":"should_call","user_input":"直接查询 eval_oss_volume 里的 CSV 文件，看前 10 行","expected_skill":"clickzetta-volume-manager","expected_output_contains":["eval_oss_volume"]}
+{"case_id":"005","type":"should_call","user_input":"删除 eval_oss_volume 和 eval_oss_conn，清理测试资源","expected_skill":"clickzetta-volume-manager","expected_output_contains":["eval_oss"]}

package/bin/skills/cz-cli-inner/SKILL.md

@@ -37,8 +37,8 @@ cz-cli task save-content <task> --file <f>    Save task script
 cz-cli task save-config <task>                Save task non-cron config, like retry, dependency
 cz-cli task save-cron <task>                  Save task schedule config
 cz-cli task deps <task>                       Show task dependencies (draft)
-cz-cli task online <task>                     Publish a task
-cz-cli task offline <task>                    Take task offline (irreversible)
+cz-cli task deploy <task>                     Publish/deploy a task (alias: online)
+cz-cli task undeploy <task>                   Undeploy a task, irreversible (alias: offline)
 cz-cli task execute <task>                    Execute ad-hoc
 cz-cli task delete <task>                     Delete draft/offline task
 cz-cli task flow dag <task>                   Get flow DAG
@@ -78,6 +78,7 @@ cz-cli datasource objects <name_or_id> <catalog>
                                               List objects (tables/topics/collections) in a catalog
 cz-cli datasource describe <name_or_id> <catalog> <object>
                                               Show object metadata (columns, types)
+cz-cli datasource test <name_or_id>          Test data source connectivity
 ```
 
 ## Output Formats
@@ -92,9 +93,9 @@ cz-cli datasource describe <name_or_id> <catalog> <object>
 1. **SQL is async by default**. Use `--sync` for SELECT when you need data immediately.
 2. **Write operations require `--write` flag** (INSERT/UPDATE/DELETE/CREATE/DROP).
 3. **Always pass `--type` when creating tasks** (SQL/PYTHON/SHELL/SPARK/FLOW).
-4. **Flow tasks use `task flow *` commands exclusively** — never use `task save-content` or `task online` on flow nodes.
+4. **Flow tasks use `task flow *` commands exclusively** — never use `task save-content` or `task deploy` on flow nodes.
 5. **Paginated results**: `list` commands return page 1 only. Check `ai_message` in response for next-page hints.
-6. **State-changing operations** (online/offline/execute/delete/refill): confirm intent with user first.
+6. **State-changing operations** (deploy/undeploy/execute/delete/refill): confirm intent with user first.
 7. **Multi-environment**: use `--profile <name>` to target a specific environment.
 8. **On `NO_PROFILE` error**: guide user to run `cz-cli setup`.
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@clickzetta/cz-cli-darwin-arm64",
-  "version": "0.3.20",
+  "version": "0.3.21",
   "description": "cz-cli binary for macOS ARM64 (Apple Silicon)",
   "os": ["darwin"],
   "cpu": ["arm64"],

package/bin/skills/clickzetta-data-ingest-pipeline/SKILL.md

@@ -1,220 +0,0 @@
----
-name: clickzetta-data-ingest-pipeline
-description: |
-  ClickZetta Lakehouse 数据导入总览与路由。根据用户的数据源类型、实时性要求、数据量等条件，
-  推荐最合适的数据导入方式，并引导到对应的专项 Skill 或直接执行简单导入操作。
-  当用户说"导入数据到 Lakehouse"、"数据入仓"、"数据入湖"、"怎么把数据导进来"、
-  "数据采集"、"数据加载"、"ingest data"、"load data"、"数据导入方案选择"时触发。
-  Keywords: data ingestion, import, routing, pipeline selection, data source
----
-
-# Lakehouse 数据导入总览与路由
-
-根据用户的数据源、实时性需求、数据规模等条件，推荐最合适的数据导入方式，
-并路由到对应的专项 Pipeline Skill 或直接执行简单导入操作。
-
-## 适用场景
-
-- 用户想把数据导入 ClickZetta Lakehouse，但不确定用哪种方式
-- 用户描述了数据源（Kafka、MySQL、OSS、文件等），需要推荐导入方案
-- 用户需要了解各种导入方式的适用场景和差异
-- 关键词：数据导入、数据入仓、数据入湖、数据采集、数据加载、pipeline 选择
-
-## 前置依赖
-
-- ClickZetta Lakehouse 账户，具备创建工作空间、Schema、表、PIPE、任务等权限
-- **执行环境（满足其一即可，优先使用 cz-cli）**：
-  - **cz-cli 路径**：已安装 cz-cli（`pip install cz-cli`），并完成 `cz-cli configure` 配置
-  - **MCP 路径**：clickzetta-studio-mcp 或 clickzetta-mcp-server 工具可用（`LH_execute_query`、`create_task`、`save_integration_task` 等）
-
-## 环境探测（执行前必读）
-
-在开始任何操作前，先判断当前执行环境：
-
-**第一步：检测 cz-cli 是否可用**
-```bash
-cz-cli --version
-```
-- 若命令存在 → **走 cz-cli 路径**（见本文档末尾"cz-cli 替代路径"章节，以及各专项 Skill 的 cz-cli 替代路径）
-- 若命令不存在 → 继续检测 MCP
-
-**第二步：检测 MCP 是否可用（仅在 cz-cli 不可用时）**
-
-尝试调用 `LH_execute_query` 工具执行一条简单 SQL（如 `SELECT 1`）。
-- 若工具存在于 tool list → **走 MCP 路径**（本文档默认路径）
-- 若工具不存在 → 停止执行，提示用户：
-  > "当前环境既无 cz-cli 也无 MCP 工具，请安装其中之一后重试。
-  > cz-cli 安装：`pip install cz-cli`，然后运行 `cz-cli configure`
-  > MCP 安装：参考 clickzetta-studio-mcp 或 clickzetta-mcp-server 配置文档"
-
-## 数据导入方式决策树
-
-### 步骤 1：确认数据源类型和需求
-
-向用户收集以下信息：
-
-1. **数据源类型**：Kafka / 对象存储(OSS/S3/COS) / 关系型数据库(MySQL/PostgreSQL/SQL Server) / 本地文件 / URL/Web 文件 / Java SDK / ZettaPark
-2. **实时性要求**：实时（秒级延迟）/ 准实时（分钟级）/ 离线批量（小时/天级）
-3. **同步范围**：单表 / 多表 / 整库
-4. **是否需要持续同步**：一次性导入 / 持续增量同步
-5. **是否需要 CDC（变更数据捕获）**：是 / 否
-
-### 步骤 2：根据决策矩阵推荐方案
-
-| 数据源 | 实时性 | 同步范围 | 推荐方式 | 对应 Skill |
-|--------|--------|---------|---------|-----------|
-| Kafka | 实时/准实时 | 单 topic | Kafka PIPE 持续导入（SQL） | `clickzetta-kafka-ingest-pipeline` |
-| Kafka | 实时 | 多 topic | Studio 实时同步 | `clickzetta-realtime-sync-pipeline` |
-| 对象存储 (OSS/S3/COS) | 准实时/批量 | 文件持续到达 | PIPE 持续导入 | `clickzetta-oss-ingest-pipeline` |
-| 对象存储 | 一次性 | 批量文件 | COPY INTO 命令 | `clickzetta-file-import-pipeline`（COPY INTO 部分） |
-| MySQL/PostgreSQL/SQL Server | 实时 CDC | 单表 | Studio 实时同步 | `clickzetta-realtime-sync-pipeline` |
-| MySQL/PostgreSQL/SQL Server | 实时 CDC | 多表/整库 | Studio 多表实时同步 | `clickzetta-cdc-sync-pipeline` |
-| MySQL/PostgreSQL/SQL Server | 离线批量 | 单表 | Studio 离线同步 | `clickzetta-batch-sync-pipeline` |
-| MySQL/PostgreSQL/SQL Server | 离线批量 | 多表 | Studio 多表离线同步 | `clickzetta-batch-sync-pipeline` |
-| 本地文件 / URL | 一次性 | 单文件/多文件 | URL 下载 + COPY INTO | `clickzetta-file-import-pipeline` |
-| 流式增量计算 | 准实时 | 表变更驱动 | Dynamic Table + Stream | `clickzetta-incremental-compute-pipeline` |
-| Java 应用 | 实时/批量 | 程序写入 | Java SDK | （见下方 SDK 导入指引） |
-| Python/ZettaPark | 批量 | DataFrame | ZettaPark save_as_table | （见下方 SDK 导入指引） |
-
-### 步骤 3：路由到专项 Skill 或直接执行
-
-根据推荐方案，执行以下路由逻辑：
-
-**有对应专项 Skill 的场景** → 告知用户推荐方案，引导使用对应 Skill：
-- `clickzetta-kafka-ingest-pipeline`：Kafka PIPE 管道搭建
-- `clickzetta-oss-ingest-pipeline`：对象存储 PIPE 管道搭建
-- `clickzetta-batch-sync-pipeline`：Studio 离线同步任务
-- `clickzetta-realtime-sync-pipeline`：Studio 实时同步任务
-- `clickzetta-cdc-sync-pipeline`：Studio 多表实时同步（CDC）
-- `clickzetta-incremental-compute-pipeline`：Dynamic Table + Stream 增量计算管道
-- `clickzetta-file-import-pipeline`：URL/文件下载导入
-- `clickzetta-table-stream-pipeline`：Table Stream 变更数据捕获
-
-**无专项 Skill 的简单场景** → 直接执行：
-
-#### SQL INSERT 导入（小数据量）
-```sql
--- 使用 LH_execute_query 执行
-INSERT INTO schema_name.table_name (col1, col2, col3)
-VALUES ('val1', 'val2', 'val3');
-```
-
-#### COPY INTO 快速导入（从 Volume）
-```sql
--- 1. 确认 Volume 中有文件
-SHOW VOLUME DIRECTORY volume_name;
-
--- 2. 执行 COPY INTO
-COPY INTO schema_name.table_name
-FROM VOLUME volume_name
-USING CSV
-OPTIONS('header' = 'true');
-```
-
-#### Java SDK 导入指引
-提供 Java SDK 的关键配置信息：
-- Maven 依赖坐标
-- 连接配置（endpoint、workspace、schema、vcluster）
-- 批量写入 API：`BulkloadWriter`
-- 实时写入 API：`RealtimeWriter`
-- 建议用户参考官方文档：`comprehensive_guide_to_ingesting_javasdk_buckload_realtime`
-
-#### ZettaPark (Python) 导入指引
-- `INSERT` 方式：`session.sql("INSERT INTO ...")`
-- `save_as_table` 方式：`df.write.save_as_table("table_name")`
-- 建议用户参考官方文档：`comprehensive_guide_to_ingesting_zettapark_save_as_table`
-
-## 数据入仓 vs 数据入湖
-
-| 维度 | 数据入仓 | 数据入湖 |
-|------|---------|---------|
-| 目标 | Lakehouse 托管表 | 用户 Volume（对象存储） |
-| 格式 | 自动转为内部列式格式 | 保持原始文件格式 |
-| 查询性能 | 高（列式存储 + 索引） | 较低（需扫描原始文件） |
-| 适用场景 | 分析查询、BI 报表、数据仓库 | 数据暂存、原始数据归档、跨系统共享 |
-| 常用方式 | Studio 同步、PIPE、COPY INTO、SDK | PUT 文件、Python 脚本上传 |
-
-## 示例
-
-### 示例 1：用户不确定导入方式
-
-用户说："我有一个 MySQL 数据库，想把里面的订单表实时同步到 Lakehouse"
-
-路由逻辑：
-1. 数据源：MySQL（关系型数据库）
-2. 实时性：实时
-3. 同步范围：单表
-4. 需要 CDC：是（实时同步意味着需要捕获变更）
-→ 推荐：Studio 实时同步
-→ 路由到 `clickzetta-realtime-sync-pipeline` Skill
-
-### 示例 2：多种数据源混合场景
-
-用户说："我们有 Kafka 的用户行为日志，还有 MySQL 的业务数据，都要导入 Lakehouse"
-
-路由逻辑：
-1. Kafka 用户行为日志 → `clickzetta-kafka-ingest-pipeline`（PIPE 持续导入）
-2. MySQL 业务数据 → 确认实时性需求：
-   - 实时 → `clickzetta-realtime-sync-pipeline` 或 `clickzetta-cdc-sync-pipeline`
-   - 离线 → `clickzetta-batch-sync-pipeline`
-→ 分别引导到对应 Skill
-
-### 示例 3：简单的一次性文件导入
-
-用户说："我有一个 CSV 文件要导入"
-
-路由逻辑：
-1. 数据源：本地文件
-2. 一次性导入
-→ 路由到 `clickzetta-file-import-pipeline` Skill（支持文件上传 + COPY INTO）
-
-## 错误处理
-
-| 场景 | 处理方式 |
-|------|---------|
-| 用户无法确定数据源类型 | 询问数据当前存储位置（哪个系统/服务），帮助判断 |
-| 用户需求跨多种导入方式 | 拆分为多个独立的导入任务，分别路由到对应 Skill |
-| 推荐的 Skill 尚未创建 | 提供该导入方式的基本步骤和关键 SQL/API，引导用户参考官方文档 |
-| 用户的云环境不支持某种连接 | 使用 `LH_show_object_list`（object_type=CONNECTIONS）检查可用连接类型，推荐替代方案 |
-| 数据量极大（TB 级） | 建议分批导入，优先使用 PIPE 或 Studio 同步任务（支持断点续传） |
-
-## 注意事项
-
-- 本 Skill 是路由入口，不直接执行复杂的 pipeline 搭建，而是引导到专项 Skill
-- 对于简单场景（SQL INSERT、单次 COPY INTO），可以直接在本 Skill 中完成
-- 推荐方案时需考虑用户的云环境（阿里云/腾讯云/AWS），不同环境支持的连接类型可能不同
-- 使用 `LH_show_object_list`（object_type=VCLUSTERS）确认可用的虚拟集群，同步任务需要 SYNC 类型的 VCluster
-- 数据入仓是最常见的场景，数据入湖主要用于原始数据暂存或跨系统共享
-
----
-
-## cz-cli 替代路径
-
-> 仅在 cz-cli 可用且 MCP 不可用时使用本节。
-> 本 Skill 是路由入口，cz-cli 路径的核心逻辑在各专项 Skill 的"cz-cli 替代路径"章节中。
-
-### 路由说明
-
-当 MCP 不可用时，各专项 Skill 均已提供 cz-cli 替代路径：
-
-| 数据源 | 推荐方式 | 对应 Skill 的 cz-cli 路径 |
-|--------|---------|--------------------------|
-| Kafka | PIPE 持续导入 | `clickzetta-kafka-ingest-pipeline` → cz-cli 替代路径 |
-| 对象存储 (OSS/S3/COS) | PIPE 持续导入 | `clickzetta-oss-ingest-pipeline` → cz-cli 替代路径 |
-| MySQL/PostgreSQL/SQL Server（实时单表） | Studio 实时同步 | `clickzetta-realtime-sync-pipeline` → cz-cli 替代路径 |
-| MySQL/PostgreSQL/SQL Server（实时多表/整库） | Studio 多表实时同步 | `clickzetta-cdc-sync-pipeline` → cz-cli 替代路径 |
-| MySQL/PostgreSQL/SQL Server（离线批量） | Studio 离线同步 | `clickzetta-batch-sync-pipeline` → cz-cli 替代路径 |
-
-### 简单场景直接执行（cz-cli 版）
-
-对于无需专项 Skill 的简单场景，可直接用 cz-cli agent 完成：
-
-```bash
-# SQL INSERT 导入（小数据量）
-cz-cli agent run "向表 <schema_name>.<table_name> 插入数据：<col1>=<val1>, <col2>=<val2>" \
-  --format a2a --dangerously-skip-permissions
-
-# COPY INTO 快速导入（从 Volume）
-cz-cli agent run "从 Volume <volume_name> 以 CSV 格式（有 header）将数据导入表 <schema_name>.<table_name>" \
-  --format a2a --dangerously-skip-permissions
-```

package/bin/skills/clickzetta-data-ingest-pipeline/eval_cases.jsonl

@@ -1,5 +0,0 @@
-{"case_id":"001","type":"should_call","user_input":"我想把数据导入 Lakehouse，但不确定用哪种方式","expected_skill":"clickzetta-data-ingest-pipeline","expected_output_contains":["数据源","实时","批量"]}
-{"case_id":"002","type":"should_call","user_input":"数据入仓有哪些方案？怎么选择？","expected_skill":"clickzetta-data-ingest-pipeline","expected_output_contains":["Kafka","对象存储","MySQL"]}
-{"case_id":"003","type":"should_call","user_input":"我有 MySQL 和 Kafka 两个数据源要导入 Lakehouse，分别用什么方式？","expected_skill":"clickzetta-data-ingest-pipeline","expected_output_contains":["CDC"]}
-{"case_id":"004","type":"should_call","user_input":"数据导入方案怎么选？实时和离线有什么区别？","expected_skill":"clickzetta-data-ingest-pipeline","expected_output_contains":["实时","离线","延迟"]}
-{"case_id":"005","type":"should_call","user_input":"ingest data into ClickZetta Lakehouse, what options do I have?","expected_skill":"clickzetta-data-ingest-pipeline","expected_output_contains":["Kafka","OSS","SDK"]}