Add Couchbase MCP Server llms.txt - #240
Conversation
An emerging convention that some AI crawlers and agent frameworks check for a concise, machine-readable summary of a project.
There was a problem hiding this comment.
Pull request overview
Note
Copilot was unable to run its full agentic suite in this review.
Adds an llms.txt documentation file to provide LLMs/AI assistants a concise, structured project summary and key operational facts for the Couchbase MCP Server.
Changes:
- Introduces
llms.txtwith an overview of the Couchbase MCP Server and its capabilities - Adds links to docs/README/CONTRIBUTING plus distribution references (PyPI/Docker)
- Documents key agent-facing facts (registry name, runtime, safety defaults, client support)
💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
| @@ -0,0 +1,24 @@ | |||
| # Couchbase MCP Server | |||
|
|
|||
| > Couchbase MCP Server is a self-hosted Model Context Protocol (MCP) server that lets AI agents and LLM-powered assistants (Claude, Cursor, Windsurf, VS Code Copilot, JetBrains AI Assistant/Junie, and other MCP clients) connect to and interact with data in Couchbase clusters — whether hosted on Couchbase Capella or self-managed. It exposes tools for cluster health, data schema discovery, key-value document operations, SQL++ querying, and query performance analysis, with read-only mode enabled by default for safety. | |||
There was a problem hiding this comment.
Wow Microsoft REALLY cares about its products branding huh?
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
There was a problem hiding this comment.
The file looks good, but could you give some examples of popular projects where this convention of having a llms.txt file is followed? I just want to gauge how others do it.
There was a problem hiding this comment.
It's an emerging standard to be sure, but here are a few examples:
| @@ -0,0 +1,24 @@ | |||
| # Couchbase MCP Server | |||
|
|
|||
| > Couchbase MCP Server is a self-hosted Model Context Protocol (MCP) server that lets AI agents and LLM-powered assistants (Claude, Cursor, Windsurf, VS Code Copilot, JetBrains AI Assistant/Junie, and other MCP clients) connect to and interact with data in Couchbase clusters — whether hosted on Couchbase Capella or self-managed. It exposes tools for cluster health, data schema discovery, key-value document operations, SQL++ querying, and query performance analysis, with read-only mode enabled by default for safety. | |||
There was a problem hiding this comment.
Wow Microsoft REALLY cares about its products branding huh?
|
|
||
| ## Docs | ||
|
|
||
| - [Full documentation](https://mcp-server.couchbase.com/): configuration reference, transports, OAuth, logging, and per-client setup guides |
There was a problem hiding this comment.
MCP server documentation is moved to the official documentation. Please update this.
https://docs.couchbase.com/mcp-server/get-started/overview.html
| - Default safety posture: `CB_MCP_READ_ONLY_MODE=true` — write tools are not loaded unless explicitly disabled | ||
| - Works with both Couchbase Capella (managed) and self-managed Couchbase Server | ||
| - Supports Claude Desktop, Cursor, Windsurf, GitHub Copilot (in VS Code), and JetBrains AI Assistant/Junie as MCP clients | ||
| - Community-maintained by Couchbase, not covered by the standard Couchbase support portal |
There was a problem hiding this comment.
Couchbase MCP server is transitioned to Enterprise supported starting from v1.0. It is not community supported anymore. However, the AIDP license is mandatory for the enterprise support.
Updated information regarding Couchbase MCP server support.
VirajAgarwal-couchbase
left a comment
There was a problem hiding this comment.
Thanks for the PR! And sorry again for my delayed review.
An emerging convention that some AI crawlers and agent frameworks check for a concise, machine-readable summary
Related Issue
n/a
Resolves:
What does this change do?
Improves GEO
Why is this change needed?
discoverability for LLMs
Evidence of Testing
n/a
Automated tests — commands run and results summary:
n/a
Environments tested (both are required):
n/a
Manual verification — MCP client used (Claude Desktop, Cursor, MCP Inspector, ...) and what was exercised. Screenshots or tool-call transcripts are very helpful:
local VSCode editing
Compatibility Considerations
n/a
Checklist
n/a
cb_mcp.corecontracts / managed MCP interfaces (or discussed first)