I recently encountered a product that appears to be tailor-made for AI-related applications. Ghost, developed by ghost.build, positions itself as “the first database built for agents.”
Ghost is an “agent-first” Postgres database platform designed to empower developers and AI agents to create, fork, inspect, query, manipulate, and delete entire databases effortlessly. Remarkably, it is completely free to use.
The creators of Ghost operated under a straightforward premise: if agents are tasked with building software, testing migrations, inspecting schemas, running SQL, and experimenting with data, they require databases that are as disposable and programmable as code sandboxes. Ghost fulfills this requirement seamlessly.
When a user creates a Ghost database or forks an existing one, these databases are hosted on Ghost’s Cloud infrastructure, rather than on local machines. This makes Ghost particularly advantageous for testing, prototyping, agent workflows, branch databases, migration experiments, and disposable database environments.
In contrast to traditional managed databases, which are structured around long-lived production infrastructure, Ghost allows for a more dynamic approach. Users can create databases on demand, fork them for isolated copies, execute SQL queries, inspect schemas, experiment with different table configurations, and discard them when finished.
This flexibility makes Ghost an ideal companion for AI tools such as Codex and Claude Code. These tools can reason about code, write migrations, debug queries, generate seed data, inspect logs, and utilize MCP tools. Ghost’s integrated MCP server provides Codex with direct database management capabilities, eliminating the need for vague instructions, copied connection strings, or manual dashboard operations.
N.B. I have no affiliation or association with the company or team behind Ghost or the Ghost product itself.
Prerequisites
Before diving into the installation process, ensure you have one of Ghost’s supported coding agents installed on your system. These include:
Claude Code
Codex
Cursor
Gemini CLI
Google Antigravity
Kiro CLI
VS Code
WindsurfAdditionally, a GitHub account is necessary, as you will need to grant Ghost access during the login process.
Installing Ghost
For users on Linux, WSL on Windows, or macOS, the installation can be initiated with the following curl command:
$ curl -fsSL https://install.ghost.build | shIf you are using Windows, execute this command in a PowerShell terminal:
PS C:Usersthoma> irm https://install.ghost.build/install.ps1 | iexThen, log in by running:
PS C:Usersthoma> ghost login
Opening browser for authentication...
Found space: nj5scy2orp
Successfully logged in as [email protected]Upon successful login, you will see a screen prompting you to authorize Ghost to proceed.
After logging in, the next step is to make the Ghost MCP server accessible to your preferred coding agent. For instance, if you are using Codex, enter the following command and select your agent from the list:
PS C:Usersthoma> ghost mcp install
Select an MCP client to configure:
1. Claude Code
> 2. Codex
3. Cursor
4. Gemini CLI
5. Google Antigravity
6. Kiro CLI
7. VS Code
8. Windsurf
Typing: 2
Successfully installed Ghost MCP server configuration for codex
Configuration file: C:Usersthoma.codexconfig.toml
Next steps:
1. Restart codex to load the new configuration
2. The Ghost MCP server will be available as 'ghost'
PS C:Usersthoma>Check that your setup is OK
- To verify that the Ghost MCP is installed correctly, enter the following command:
PS C:Usersthoma> ghost mcp list
TYPE NAME
prompt design-postgis-tables
prompt design-postgres-tables
prompt find-hypertable-candidates
prompt migrate-postgres-tables-to-hypertables
prompt pgvector-semantic-search
prompt postgres
prompt postgres-hybrid-text-search
prompt setup-timescaledb-hypertables
tool ghost_connect
tool ghost_create
tool ghost_create_dedicated
tool ghost_delete
tool ghost_feedback
tool ghost_fork
tool ghost_fork_dedicated
tool ghost_list
tool ghost_login
tool ghost_logs
tool ghost_password
tool ghost_pause
tool ghost_rename
tool ghost_resume
tool ghost_schema
tool ghost_sql
tool ghost_status
tool search_docs
tool view_skillNext, confirm that your agent can utilize the Ghost MCP server. You should see output similar to this (for Codex):
PS C:Usersthoma> codex mcp list
Name Command Args Env Cwd Status Auth
ghost C:UsersthomaAppDataLocalProgramsGhostghost.exe mcp start - - enabled UnsupportedIf everything checks out, we can proceed to explore practical examples of using Ghost via its CLI and agents through the MCP server.
Example 1 — Creating a sales data database
Open your coding agent and input the following command:
Codex> Create a ghost database called sales_data that stores 10,000 dummy sales
records. In the database, create whichever tables are suitable for such
data. Do not include any indexes, primary or foreign keys for now.The output from Codex was as follows:
Worked for 1m 32s
Created a fresh Ghost database named sales_data.
It contains 10,000 dummy sales order records and supporting tables:
customers 1,000
products 150
sales_reps 30
sales_orders 10,000
sales_order_items 30,000
Per your instruction, I did not create any:
primary keys
foreign keys
indexes
Verification confirmed there are no primary key constraints, no foreign key
constraints, and no indexes on the sales tables.This process took just over a minute and a half, showcasing the efficiency of Ghost.
Example 2 — Using the Ghost CLI
Now that we have a database at our disposal, let’s explore some Ghost CLI commands. While these commands are typically unnecessary since Codex and the Ghost MCP can handle most tasks, it’s beneficial to understand their functionality.
- To list the databases, enter:
PS C:Usersthoma> ghost list
ID NAME STATUS STORAGE
sip3qcs40a sales_data running 211MiB- To view the database tables and structure, use:
$ curl -fsSL https://install.ghost.build | sh0
- To select data from the sales_order table, execute:
$ curl -fsSL https://install.ghost.build | sh1
- Concerned about your agent potentially damaging your database? You can easily set your system to read-only:
$ curl -fsSL https://install.ghost.build | sh2
A plethora of useful commands are available with the Ghost CLI. For a comprehensive list, simply type ghost — help into your command line.
However, as previously mentioned, one of the primary advantages of Ghost is the ability to manipulate databases using an AI agent. Let’s demonstrate this capability in action.
Example 3 — Using the Ghost MCP with our agent
We previously observed how swiftly a database can be created using Codex. Now, let’s illustrate how easy it is to experiment with databases using an agent.
- To add 500,000 records to the sales_order_items table, input:
$ curl -fsSL https://install.ghost.build | sh3
- To select data from the sales_order_id table, use:
$ curl -fsSL https://install.ghost.build | sh4
- For performance tuning, execute:
$ curl -fsSL https://install.ghost.build | sh5
The performance improvement after applying tuning was significant, showcasing a remarkable +200x enhancement in query runtime.
Example 4 — Building a dashboard app for our database
This example illustrates the seamless integration of an app developed with coding agents and a database constructed with the same tools.
$ curl -fsSL https://install.ghost.build | sh6
Upon clicking the HTML link generated, I was greeted with a dynamic dashboard that accurately responded to various filter values.
Summary
Through these examples, the capabilities of Ghost’s database have been effectively showcased, highlighting its potential for rapid database manipulation and experimentation.
One of the standout features of Ghost is the ease with which users can conduct what-if scenarios with real databases. Whether you need to create multiple copies of an existing database for testing or modify indexes and primary keys, Ghost simplifies the process through natural language commands with your preferred coding agents, while still offering CLI options for more complex needs.
As demonstrated in the final example, Ghost’s compatibility with agentic coding tools facilitates the development of applications that integrate seamlessly with Ghost databases, enabling the creation of useful tools such as dashboards and other SaaS applications.
While it may be prudent to exercise caution before storing production data on Ghost, it serves as an excellent platform for proofs of concept and general experimentation. Once a satisfactory database structure is established using Ghost, replicating that structure in a production environment should be straightforward.
For further information on utilizing Ghost and accessing comprehensive documentation, please visit the official homepage: https://ghost.build/docs/#introduction.
