LangGraph RAG MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| ANTHROPIC_API_KEY | Yes | Anthropic API key for Claude models |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| langgraph_query_toolC | Search local LangGraph docs; return excerpts, source URLs and cosine similarity. |
| read_pageA | Read an indexed source URL from search results, with character-based pagination. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| index_status | Return index timestamp, source, model revision and document/chunk counts. |
| full_docs | Read the corpus if it fits the 200,000-character limit; prefer search/read_page. |
TDQS
Scored across 2 tools
langgraph_query_tool and read_page have clearly distinct purposes: one searches indexed docs and returns excerpts/URLs, the other reads a specific URL with pagination. No overlap; the intended workflow is search then read.
Both use snake_case, but langgraph_query_tool includes a domain prefix and redundant _tool suffix while read_page is a simple verb_noun. The inconsistency in naming pattern makes it mixed but still readable.
Two tools is thin for a RAG server; a typical retrieval interface might include a list-sources or metadata tool. The pair covers core search-and-read but feels minimal.
The search and read tools cover the essential retrieval lifecycle (query -> fetch full content). Minor gaps exist, such as listing all indexed documents or retrieving metadata, but agents can work around them.