Graph RAG MCP Server for Obsidian (saselvan/graph-rag-mcp-server-fixed) is an MCP server listed on the M8ven Trust Index. It scores 52 out of 100, grade D. It declares 40 tools. No publisher has claimed this listing.

Caution

52/100

Graph RAG MCP Server for Obsidian

Item: Graph RAG MCP Server for Obsidian
Rating: 52
Author: M8ven

A local-first Graph-RAG system combining ChromaDB with metadata-based graph relationships and Gemini 2.5 Flash for intelligent Q&A over Obsidian vaults, supporting MCP clients like Claude Desktop, Cursor, and Raycast.

→ saselvan/graph-rag-mcp-server-fixed · listed on glama

Caution. Specific findings reduced this grade. They are listed on the page. Grades reflect the full trust pyramid: code, verification depth, and reputation. New projects cap at C until adoption is earned.

How we verified

⚡ Live Monitored: not connected

Verified is a snapshot. Live keeps it current, and builds your track record.

⚡ Connect GitHub → continuous verification on every push why connect →

Who stands behind it

saselvan

Source: Glama

Is this your MCP?

Claim it to get a verified publisher badge, a free copy of our full audit findings, and direct contact for any high-priority issues we find. Or connect your repo for our deepest verification, Live Monitored: read-only, revoke anytime. What we access →

⚡ Connect GitHub for Live Monitored

Install from

Glamanot scored here

The grade above is for the source repository. Registries can serve a different version, so we mark the ones we were not able to read.

// key findings

⚠️

Tool descriptions don’t match what handlers do

1 tool describes read intent but its handler mutates — enrich_base_with_graph (line 2960: base_path.write_text(base_json, encoding='utf-8'))

✅

No credential exfiltration, no sensitive file access, no obfuscation

Static analysis found nothing flowing your secrets to unexpected places.

🔐

You'll be asked for 1 credential: GEMINI_API_KEY

These are read from process.env at runtime. Make sure you trust where they’ll be sent.

// tools this server exposes37 tools

These names and descriptions are the publisher's own, read from the source code. We print them as written. Our assessment is the findings above, not this list.

smart_search

Intelligent search that analyzes query intent and chooses optimal search strategy. Always returns a JSON-serializable dict. On any internal failure, falls back to a minimal, empty result instead of surfacing transport-level errors. This replaces basic search with smart routing between: - Vector simi…

health_check

Run registered health checks and return component status.

get_dspy_optimization_status

Return current DSPy optimization status and scheduling details.

force_dspy_optimization

Force a DSPy optimization run if available and not already running.

search_notes

Basic vector search across vault chunks using ChromaDB. Note: Consider using smart_search instead for better results with intelligent routing.

answer_question

Enhanced RAG-powered Q&A using intelligent search routing and Gemini 2.5 Flash. Uses smart search routing to find the most relevant content through: - Vector similarity for semantic questions - Graph traversal for relationship questions - Tag matching for categorical questions - Hybrid approach for …

graph_neighbors

Get neighboring notes in the graph up to specified depth.

get_subgraph

Get a subgraph containing seed notes and their neighbors.

list_notes

List notes with optional vault and time-based filtering.

read_note

Read the full content of a note by path.

get_note_properties

Get frontmatter properties of a note.

update_note_properties

Update frontmatter properties of a note.

archive_note

Move a note to the archive folder.

create_note

create_folder

Create a new folder in the vault.

add_content_to_note

Add content to an existing note.

update_note_section

get_backlinks

Get all notes that link to the specified note.

get_notes_by_tag

Get all notes that have the specified tag.

traverse_from_chunk

Traverse the graph starting from a specific semantic chunk. This provides chunk-level graph navigation, allowing you to explore relationships from any specific piece of content in the vault.

get_related_chunks

Find chunks related to a query through graph relationships. Unlike vector search, this finds chunks connected through the graph structure (links, backlinks, hierarchical relationships).

explore_chunk_context

Get surrounding context for a specific chunk. Returns the chunk along with its sequential neighbors and hierarchical context, useful for understanding the full context around a specific piece of content.

reindex_vault

enrich_notes

Enrich notes with PARA taxonomy and semantic relationships. Args: note_paths: Specific note paths to enrich. If None, enriches all notes. limit: Maximum number of notes to process (if note_paths is None) dry_run: If True, analyze but don't save changes (default: False) Returns: EnrichmentResult with…

list_bases

List all .base files in the vault with their metadata. Returns: List of BaseInfo objects containing base file information

read_base

Read and parse a .base file by its ID. Args: base_id: The base file ID to read Returns: Parsed base file configuration as dictionary, or None if not found

validate_base

Validate .base file content syntax and structure. Args: content: The .base file content to validate format: Optional format hint (json or yaml) Returns: Dictionary with 'valid' boolean and optional 'error' message

execute_base_query

Execute a base query and return matching notes. Args: base_id: The base file ID to execute view_id: Optional specific view to use limit: Optional limit on number of results Returns: Query results with matching notes and metadata

get_base_view

Get formatted data for a specific base view. Args: base_id: The base file ID view_id: The view ID to format Returns: Formatted view data for display

create_base

Create a new .base file with basic configuration. Args: name: Name for the new base description: Optional description folders: Folders to include (defaults to root) filters: Optional filter conditions Returns: Dictionary with 'success' boolean, 'base_id', 'path', and optional 'error'

list_base_files

List all .base files in the vault. Returns: Dictionary with 'success' boolean, 'base_files' list, and optional 'error'

check_base_exists

Check if a base file exists. Args: base_id: The base file ID to check Returns: Dictionary with 'exists' boolean, file info if exists, and optional 'error'

update_base

Update an existing .base file configuration. Args: base_id: The base file ID to update updates: Dictionary of updates to apply Returns: Dictionary with 'success' boolean and optional 'error'

base_from_graph

Create a new .base file from a selection of notes (graph-friendly). This tool enables creating bases from graph selections, making it easy to save curated collections of related notes discovered through graph traversal. Args: note_ids: List of note IDs to include name: Name for the new base descript…

enrich_base_with_graph

Enrich a base with graph relationship data as computed fields. Adds computed fields for graph metrics like link count, centrality, and cluster information to enhance base views with graph insights. Args: base_id: The base file ID to enrich Returns: Dictionary with 'success' and enrichment details

get_cache_stats

Get detailed cache statistics for monitoring and optimization.

clear_caches

Clear specified caches or all caches if no names provided. Args: cache_names: Optional list of cache names to clear. If None, clears all caches. Returns: Status of cache clearing operation.

// environment variables

To run this server yourself, you supply these values. They go in your own MCP client configuration and stay on your machine. The secret label means the value is sensitive, not that the server mishandles it.

configOBSIDIAN_RAG_CHROMA_DIR/custom/path/to/.chroma_db

🔐 secretGEMINI_API_KEYEdit .env and add your and vault paths

configOBSIDIAN_VAULT_PATH

configOBSIDIAN_RAG_VAULTS"": "/path/to/your/vault"

configOBSIDIAN_RAG_VAULT_PATH

configCHROMA_DIR

configOBSIDIAN_RAG_DSPY_STATE_DIR

configDSPY_STATE_DIR

configXDG_CACHE_HOMEIntegration tests run without network. Embeddings fall back to a built‑in default and DSPy caches write to .cache/. Override with or DSPY_CACHEDIR if needed.

configDSPY_CACHEDIRIntegration tests run without network. Embeddings fall back to a built‑in default and DSPy caches write to .cache/. Override with XDG_CACHE_HOME or if needed.

// quality suggestions

Tool annotations

No tools have read-only/destructive annotations

Add readOnlyHint or destructiveHint annotations to every tool so hosts can warn users before invoking.

All four hints declared on every tool

40/40 tools missing one or more hints — smart_search (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint); health_check (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint); get_dspy_optimization_status (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint), +37 more. OpenAI's directory rejects tools where any of the four hints are missing or non-boolean.

For every tool, set all four hints (readOnlyHint, destructiveHint, idempotentHint, openWorldHint) to explicit true/false values that match the handler’s actual behaviour.

Descriptions match behaviour

1 tool describes read intent but its handler mutates — enrich_base_with_graph (line 2960: base_path.write_text(base_json, encoding='utf-8'))

Rename the tool, rewrite the description, or move the side-effect into a separate clearly-named tool.

Tool inputs are validated

31/40 tool handlers declare input schemas (78%)

Declare an inputSchema with zod/joi/yup on every tool definition.

Tool handlers catch errors

31/40 tool handlers wrap calls in try/catch (78%)

Wrap each tool handler body in try/catch and return a structured error response.

License file

No license file

Add a LICENSE file (MIT, Apache-2.0, etc.).

Tests exist

No test files found

Add tests that exercise each declared tool.

Claim the listing to review these findings one by one and send us a correction where you disagree, straight to the team. Claiming also means we tell you when the grade moves, and reach you first if we find anything urgent.

// full audit trail

The findings above are the summary. The full trail, every check we ran, each deduction, the network hosts observed and the dependency advisories, goes to verified publishers, along with an alert whenever a new one lands. Verified publishers can also review each finding and dispute it in one click. Publisher corrections have sharpened several of our checks this month, because the maintainer knows the codebase better than any scanner.

// improvement guidance — verified publishers only

We have 7 concrete improvements we can share with the publisher of this MCP. Each comes with specific guidance to raise the trust score.

// embed badge in your README

[![M8ven Verified](https://m8ven.ai/badge/mcp/saselvan/graph-rag-mcp-server-fixed?variant=verified)](https://m8ven.ai/mcp/saselvan/graph-rag-mcp-server-fixed)

Shows verification status without the grade. Want the grade badge instead? Remove ?variant=verified from the URL.

commit: 7cb9c5eab0b867e8755fd6cfe76773a8c64f52e9

code hash: 843002b980b5adf4ce2fa61d116aa51b271455806b4427edb1b9487aebd7dc61

view raw JSON →

Check MCPs from inside your assistant

Tool Check · MCP

Vetting this one by hand? Tool Check is an MCP that scores other MCPs. Add it once and ask Claude, ChatGPT, or any MCP client to grade a server, surface CVEs, check the publisher, and suggest safer alternatives — before you install.

https://m8ven.ai/api/mcp/tool-check

check_toolsearch_toolscompare_toolsrecommend_alternativescheck_publisherreport_concern

How to add it →Free · no account needed · works in any MCP client