Databricks Field Guide by TechFabric
Interactive Databricks learning, certification practice and sourced engineering guidance.
Installation
Ajoutez ceci à la configuration MCP de votre client (Claude Desktop : claude_desktop_config.json ; Cursor : .cursor/mcp.json ; VS Code et les autres utilisent la même forme « mcpServers »).
claude mcp add --transport http databricks-field-guide-by-techfabric 'https://databricks.techfabric.com/mcp'
{
"mcpServers": {
"databricks-field-guide-by-techfabric": {
"type": "http",
"url": "https://databricks.techfabric.com/mcp"
}
}
}
Documentation
Tirée du README du projet sur GitHub, actualisée chaque semaine. Les liens ouvrent GitHub.
Sommaire
Databricks Field Guide MCP server
Coding agents write Databricks code from training data that's months old. The Databricks Field Guide is what TechFabric's engineers know about building, running and paying for Databricks, and its MCP server hands your agent the section that answers the question, with code that runs and the docs.databricks.com pages that section cites.
https://databricks.techfabric.com/mcp
It's a remote server over Streamable HTTP, so there's nothing to install. It's read-only, it needs no sign-in, and it's free.
Ask how to add a Unity Catalog row filter, and the agent gets the guide's section on it, the SQL function from that section and the Databricks documentation behind it. Databricks' own docs stay the authority on syntax and limits, which is why every answer links to them.
This isn't Databricks' own MCP server. Databricks runs managed MCP servers inside a workspace that act on your data. This one is built and run by TechFabric, a Databricks partner, and it never connects to a workspace.
What's in this repository
Descriptor files, so MCP directories have a repository to list. The official MCP Registry entry names this repository, and GitHub's MCP Registry shows this README in Copilot. The server's code isn't here and isn't open source.
| File | What it is |
|---|---|
server.json | The server's entry in the official MCP Registry, copied as published |
glama.json | Maintainer list for Glama |
icon.png, icon-light.png | The TechFabric mark at 256 px, for dark and light backgrounds |
LICENSE | Apache-2.0, for these files only |
The live server and the registry are the source of truth. If this README and
the server's tools/list ever disagree, the server is right. A weekly GitHub
Action compares server.json with the registry and opens a pull request when
the entry has changed.
Add it to your agent
Claude Code
claude mcp add --transport http databricks-field-guide https://databricks.techfabric.com/mcp
Add --scope user to have it in every project.
Claude Desktop and claude.ai
Open Customize, then Connectors, choose Add and then Add custom
connector. Name it, paste https://databricks.techfabric.com/mcp, keep the
authentication settings Claude detects, and choose Add. On a Team or
Enterprise plan, an owner may have to add the connector for the organisation
first.
ChatGPT
At chatgpt.com/plugins, choose + and then
Add custom MCP server. Paste the URL under Connection, choose no
authentication, and create it. In a chat, type @ and pick it. If Add custom
MCP server isn't there, your workspace's admin has to turn it on.
Codex
codex mcp add databricks-field-guide --url https://databricks.techfabric.com/mcp
Or by hand, in ~/.codex/config.toml:
[mcp_servers.databricks-field-guide]
url = "https://databricks.techfabric.com/mcp"
Cursor
In ~/.cursor/mcp.json for every project, or .cursor/mcp.json for one:
{
"mcpServers": {
"databricks-field-guide": { "url": "https://databricks.techfabric.com/mcp" }
}
}
VS Code
In .vscode/mcp.json, then start it from that file or the MCP view. The tools
show up in Copilot's agent mode.
{
"servers": {
"databricks-field-guide": { "type": "http", "url": "https://databricks.techfabric.com/mcp" }
}
}
GitHub Copilot
In VS Code, use the configuration above. Copilot CLI saves the server to
~/.copilot/mcp-config.json, for every project:
copilot mcp add --transport http databricks-field-guide https://databricks.techfabric.com/mcp
Copilot's coding agent on github.com takes it in the repository's Settings, under Copilot and MCP servers. Every tool is read-only, so it can have all of them:
{
"mcpServers": {
"databricks-field-guide": {
"type": "http",
"url": "https://databricks.techfabric.com/mcp",
"tools": ["*"]
}
}
}
Anything else
Point the client at the URL with the Streamable HTTP transport and no
authentication. A client that only speaks stdio can reach it through
mcp-remote:
npx mcp-remote https://databricks.techfabric.com/mcp
Gemini CLI and OpenCode each have their own line in the setup guide.
Tools
| Tool | What it does |
|---|---|
search_guide | Finds the chapter sections, margin notes and events that answer a Databricks question, best first, each with its URL and section anchor |
get_chapter | Reads a chapter, a margin note or one section of either as Markdown, with code, FAQs and the docs.databricks.com links it cites |
get_code_examples | Returns the SQL, Python, CLI and config blocks that match a topic, with the explanation around each one |
whats_new | Lists recent Databricks changes the guide tracks, newest first: margin notes, release-stage changes and updated chapters |
list_chapters | The table of contents, with each chapter's part, summary, dates and main sections |
review_databricks_code | Checks one SQL, Python or YAML file (a bundle, job, cluster or warehouse config) against the guide's own rules. Each finding gives the line, why, the fix, the guide's sentence with its section URL and the docs it cites, and the result lists the sections worth reading |
estimate_cost | Estimates a workload's monthly cost from the figures the guide states, as a range with its assumptions. Every number carries its source sentence, section URL, date and pricing page, and where the guide has no figure it says so instead of guessing |
All seven are read-only, and none calls a model: the review and the estimate are
deterministic checks against the guide's text, and your agent does the
reasoning. A rule applies only while the guide still says what it cites. The
rules leave out what the guide doesn't argue, such as SELECT * or collect()
on their own.
Three resources come with them: fieldguide://llms.txt, one line per page;
fieldguide://whats-new, the last 30 days of changes as JSON; and
fieldguide://page/{+path}, any published page as Markdown. Hosts that render
MCP Apps show the results as an interactive
view instead of plain text: since 1.3.0 that includes the review's findings and
the estimate as a table.
Prompts
Hosts that support MCP prompts show these as slash commands. In Claude Code
they're /mcp__databricks-field-guide__<name>.
| Prompt | What it does |
|---|---|
databricks-review | Reviews this repository's Databricks setup against the guide. Your agent collects the bundles, notebooks, SQL and compute configs, runs review_databricks_code on each and summarises the findings with links |
databricks-cost | Estimates what the repository's workloads cost with estimate_cost, with every figure's source and date and the gaps stated plainly |
whats-new | Summarises what Databricks changed lately and which of it touches this repository |
field-guide-chapter | Puts a chapter, or one section of it, into the conversation as Markdown |
field-guide-code | Finds the guide's code examples for a topic, optionally from one chapter |
Chapter and section arguments complete as you type, and so does the page
resource's path. MCP completes prompt arguments, not tool arguments, so these
prompts are where get_chapter's chapter names complete.
Notifications
On MCP 2026-07-28 a client can subscribe with subscriptions/listen to
fieldguide://whats-new or any fieldguide://page/.... When a deploy of the
guide changes a subscribed resource, such as a new margin note, a rewritten
chapter or a feature moving release stage, the server sends
notifications/resources/updated. Clients on 2025-11-25 have no stream to push
on, so they read whats_new instead.
Every answer says when its page last changed and how long a client may cache it: 10 minutes for what's new, 15 for search and an hour for chapters and code. The index is rebuilt on every deploy of the guide.
Things to ask
- "How do I add a Unity Catalog row filter? Use the Databricks Field Guide."
- "What changed in Databricks in the last two weeks?"
- "Get the Field Guide's SQL for a Unity Catalog column mask and apply it to the email column in my customers table."
- "Read the Field Guide's Lakebase chapter and tell me whether branching fits our test environments."
- "What does the Field Guide say about choosing compute to cut our Databricks bill?"
- "Review this repo's Databricks bundle against the Field Guide." Or run
/mcp__databricks-field-guide__databricks-reviewin Claude Code. - "Estimate what a Medium Databricks App costs running office hours, using the Field Guide's figures."
- "Check
pipelines/orders.sqlagainst the Field Guide and fix what it flags."
Sign-in, limits and privacy
There's no sign-in and no write tool. Each IP address gets 60 calls a minute
and 300 an hour; past that the server answers HTTP 429 with a Retry-After
header.
The server doesn't keep what you searched for or read, or the code you send
to review_databricks_code. A finding that quotes a line masks any secret on
it. The server's log line for a call names the tool and the protocol revision,
never the arguments. It stores two things: the per-IP call count, which expires
after about an hour, and one record per deploy of what changed, for
notifications. It makes no outbound request. The site's
privacy policy covers the rest.
It speaks MCP revision 2026-07-28, and clients still on 2025-11-25 are served on
the same URL. In the official registry it's
com.techfabric.databricks/field-guide.
Links
- The Databricks Field Guide
- The Field Guide MCP server's documentation: every tool, resource and limit, and the setup for each client
- The official MCP Registry entry
- The techfabric.com MCP server
- TechFabric's Databricks consulting
- Talk to an engineer
License
The files in this repository are under the Apache License 2.0, copyright Tech Fabric LLC. The Field Guide's content and the server itself aren't covered by it.
Questions fréquentes
Qu'est-ce que Databricks Field Guide by TechFabric ?
Interactive Databricks learning, certification practice and sourced engineering guidance.
Comment installer Databricks Field Guide by TechFabric ?
Il fonctionne comme serveur distant. Dans Claude Code : claude mcp add --transport http databricks-field-guide-by-techfabric 'https://databricks.techfabric.com/mcp'. Pour les autres clients, collez la configuration JSON de cette page dans leurs réglages MCP.
Faut-il une clé API pour Databricks Field Guide by TechFabric ?
Son entrée au registre n'indique aucune clé ni secret obligatoire. Certaines fonctions peuvent tout de même demander un compte auprès du service auquel il se connecte.
Databricks Field Guide by TechFabric est-il open source ?
Oui. Son code est sur GitHub sous licence Apache-2.0.
Qui maintient Databricks Field Guide by TechFabric ?
Il est publié par Fabric-Pro. AI.tn le référence sans le maintenir ni le recommander.
Issu du registre MCP officiel. Être listé n'est pas une recommandation : vérifiez le code et les permissions d'un serveur avant de lui donner accès à vos données.