Skip to main content

Agentic Astra MCP Server

Project description

Agentic Astra

Agentic Astra provides a semantic layer to connect AI Agents to Astra DB through MCP.

It is a powerful solution to translate structured and unstructured data with configurable business logic about how to use it.

Is also includes a UI for tool editing and Agentic Tool Generation

Astra MCP Server Overview

The agentic-astra that provides tools to interact with Astra DB. It is built with FastMCP and Astrapy (then Astra DB or DataStax HCD can be used as database).

The server will load the tool definitions from a collection in Astra DB or a file. The tool definitions are then transformed to a function definition that can be passed to an LLM, making it possible to use the tools provided by the MCP Server in an Agentic workflow.

When a tool is called, the server will call the appropriate method in Astra DB or DataStax HCD, converting the parameters to the appropriate filters and return the result to the MCP Client/Agent. If some embedding generations is required, the models from OpenAI or IBM Watsonx can be used for similarity search.

How to run the Astra MCP Server

Running it as MCP Server with STDIO

To run the agentic-astra as MCP Server with STDIO, you can use the following command:

uvx agentic-astra --env-file .env --tr http|sse|stdio
# or
uvx agentic-astra --astra_token <astra_token> --astra_db_name <astra_db_name>

Getting Started

You will need a Astra DB database and a Astra DB application token. You can get the token from the Astra DB console.

Also, you will need a collection in Astra DB to store the tool definitions. You can create a collection in Astra DB console.

Creating Tools

Run agentic-astra-ui to create tools and save them to the Astra DB collection.

npx agentic-astra-ui --env-file .env

Agentic-Astra-UI includes an agent for tool generation. You can use it to generate tools and save them to the Astra DB collection.

Once you have your tool, you are ready to run the MCP Server.

Running it as MCP Server with HTTP locally

To run the agentic-astra as MCP Server with STDIO, you can use the following command:

uv run agentic-astra --env-file .env -tr http

Copy the .env.example file to .env and fill the variables.

Local Development

# Install dependencies
uv sync

# Run the server
uv run agentic-astra --env-file .env --reload --log-level debug

Run from the build version

# Run the server
uv build
uvx --from ./dist/agentic_astra-0.0.5-py3-none-any.whl agentic-astra --env-file .env -tr http

Using MCP Inspector (STDIO)

npx @modelcontextprotocol/inspector uv run agentic-astra --log-level debug -tr stdio --env-file .env 

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

agentic_astra-0.0.6.tar.gz (512.5 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

agentic_astra-0.0.6-py3-none-any.whl (22.5 kB view details)

Uploaded Python 3

File details

Details for the file agentic_astra-0.0.6.tar.gz.

File metadata

  • Download URL: agentic_astra-0.0.6.tar.gz
  • Upload date:
  • Size: 512.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.12

File hashes

Hashes for agentic_astra-0.0.6.tar.gz
Algorithm Hash digest
SHA256 cc66e20ae9c36f82f21f98cff221fa5d11535e81b778ab490addcd40e6837b5b
MD5 5ef0a2dcfe11d38957a52a2ec3060518
BLAKE2b-256 ecc757e0b4c23847c78bf502e1368c70844447bb8543d4b5ad9b9a435810e82d

See more details on using hashes here.

File details

Details for the file agentic_astra-0.0.6-py3-none-any.whl.

File metadata

  • Download URL: agentic_astra-0.0.6-py3-none-any.whl
  • Upload date:
  • Size: 22.5 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.12

File hashes

Hashes for agentic_astra-0.0.6-py3-none-any.whl
Algorithm Hash digest
SHA256 25f3538e8b17f2ff6d174a7fd96ebb468618c82e19663786955c143dd5cb0db0
MD5 485efbf1395e2bc696a1a26c0bcb63f7
BLAKE2b-256 30f9a70d545e2010567f1ce820b4be6002d4b5bf35d41d12a75ab03e22b53082

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page