--- title: Common tasks with Tiger CLI and Tiger MCP | Tiger Data Docs description: Manage services and work with your data in Tiger Cloud using Tiger CLI commands or Tiger MCP natural-language prompts --- Tiger CLI and Tiger MCP let you work with Tiger Cloud in complementary ways. With Tiger CLI you run commands in a terminal, which is precise, scriptable, and repeatable. With Tiger MCP you ask your AI agent in plain language: it runs the same service and data operations, and it can also reason about your database using built-in skills. There are two levels to work at: managing your services, which has no SQL, and working with your data, which is all SQL. The examples below show the CLI way, and the equivalent Tiger MCP prompt. New to these tools? See [Tiger CLI and Tiger MCP](/learn/tiger-cli-mcp/index.md). ## Prerequisites for this page To follow these steps, you'll need: - Tiger CLI installed and your AI agent connected through Tiger MCP. See [Integrate Tiger Cloud with your AI agent](/get-started/quickstart/mcp-cli/index.md). Commands and tools can change between releases, so run `tiger --help` for the current CLI, and see the [Tiger CLI reference](/reference/tiger-cloud/tiger-cli/index.md) and [Tiger MCP reference](/reference/tiger-cloud/tiger-mcp/index.md). For agent safety, see [best practices](/build/tiger-cli-mcp/agent-best-practices/index.md). ## Manage your services Every service task is a `tiger service` command in the CLI, and the Tiger MCP has a matching tool, so you can instead just describe what you want: | Task | CLI command | Ask your agent | | -------------------------------------- | ------------------------------------------------------------------------------------ | -------------------------------------------------------------------------- | | Create a service | `tiger service create --name analytics --region us-east-1` | *"Create a time-series service called analytics in us-east-1."* | | Fork a service to test a change safely | `tiger service fork --now` | *"Fork service abc123 so I can test a schema change."* | | Resize a service | `tiger service resize --cpu 4 --memory 16` | *"Resize service abc123 to 4 CPU and 16 GB."* | | Start or stop a service | `tiger service start `, `tiger service stop ` | *"Stop service abc123."* | | Rotate the database password | `tiger service update-password ` | *"Reset the password on service abc123."* | | List, inspect, and check logs | `tiger service list`, `tiger service get `, `tiger service logs --tail 100` | *"List my services and show recent logs for abc123."* | | Delete a service | `tiger service delete ` | Not available. There is no Tiger MCP tool for delete; use the CLI command. | ## Work with your data Everything inside the database is SQL. With Tiger CLI, open a session and run it (add `--read-only` for a session that blocks writes): Terminal window ``` tiger db connect ``` With Tiger MCP, you skip the connection and describe the task, and the agent runs the SQL against the service you name. For example, to create a hypertable: ``` CREATE TABLE metrics ( time timestamptz NOT NULL, device text, value double precision ) WITH (tsdb.hypertable, tsdb.partition_column = 'time'); ``` Or ask *"Create a hypertable called metrics partitioned by time, with device and value columns."* The same pattern covers the rest of your everyday data work. Run the SQL yourself, or describe it to your agent: - Query your data, for example *"What is the hourly average value in metrics over the last day?"* - Load data from a file with `\COPY`, or ask your agent to add small amounts, for example *"Insert a few sample rows into metrics for testing."* - Roll up a continuous aggregate, for example *"Create a continuous aggregate on metrics for hourly averages per device."* - Add compression and retention, for example *"Convert metrics chunks older than 7 days to the columnstore and drop data older than 30 days."* ## Analyze, design, and optimize with Tiger MCP Some data work goes beyond running SQL. Using its built-in skills, Tiger MCP can reason about your database, which the CLI cannot do. Design and analysis are read and advice oriented, so they work in [read-only mode](/build/tiger-cli-mcp/agent-best-practices#restrict-agents-to-read-only/index.md): - Design a schema: *"Design a Tiger Cloud schema for storing IoT device telemetry."* - Find hypertable candidates: *"Analyze my database and tell me which tables should be hypertables."* - Review and optimize: *"Review my schema and indexes against best practices and suggest improvements."* Planning a change is also read-only, but testing it needs write access: forking a service (`service_fork`) is itself one of the tools read-only mode disables. - Plan a schema change: *"Plan a zero-downtime migration to add a status column to metrics, and test it on a fork first."* ## Next steps - [Cookbook](/build/tiger-cli-mcp/cookbook/index.md): Longer, multi-step prompts for schema design, data analysis, and performance tuning. - [Best practices for AI agents](/build/tiger-cli-mcp/agent-best-practices/index.md): Use read-only mode, forks, and guardrails to work safely with agents. - [Tiger CLI reference](/reference/tiger-cloud/tiger-cli/index.md): Every command, flag, and configuration parameter. - [Tiger MCP reference](/reference/tiger-cloud/tiger-mcp/index.md): Every MCP tool and its parameters. - [File an issue in the Tiger CLI repo](https://github.com/timescale/tiger-cli): Report bugs or request features to help shape the open-source Tiger CLI and Tiger MCP.