Skip to main content
Glama
Sign In
Reflex Docs MCP Server

Reflex Docs MCP Server

by itsmeadarsh2008
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Hybrid

Reflex Docs MCP Server

Version 0.2.0 — 14 tools, connection pooling, FTS5 query upgrades, TTL caching

New in 0.2.0

New Tools

  • get_code_examples(topic) — Find Python code examples from docs for a topic

  • decode_error(error_text) — Analyze a Reflex error and find relevant docs

  • get_changelog(version?) — Fetch Reflex release notes from GitHub

  • get_migration_guide(from_version, to_version) — Get migration guidance between versions

  • search_api_reference(symbol) — Look up API reference for a Reflex symbol

  • get_component_props(name) — Get only the props table for a component

  • list_recipes(category?) — Browse recipe/tutorial pages

Enhancements

  • search_docs — New fuzzy flag (prefix expansion), include_content flag (full page content)

  • get_doc — New extract_code flag (return only fenced code blocks)

Performance

  • SQLite connection pooling via threading.local() with persistent per-thread connections

  • WAL journal mode, 64 MB page cache, 256 MB mmap for concurrent reads

  • LRU cache (512 entries) on slug lookups

  • In-process TTL cache (300s default) on search results

  • FTS5 query preprocessor: phrase match + prefix expansion

  • Incremental indexing — skips re-index if git commit unchanged

Ground AI agents in real, up-to-date Reflex docs via a fast, local MCP server.

Python: 3.14 recommended. Compatible with 3.13+.

Related MCP server: LlamaIndex Documentation MCP Server

What It Does

  • Full‑text search over Reflex docs (SQLite FTS5)

  • Section‑level retrieval for precise context

  • Component index for rx.* lookups

  • FastMCP server with stdio and SSE transports

Quickstart

# Create venv
python3.14 -m venv .venv
. .venv/bin/activate

# Install (local checkout)
pip install -e .

# Index docs (clones Reflex docs and builds search index)
python -m reflex_docs_mcp.indexer

# Run MCP server (stdio)
python -m reflex_docs_mcp.server

# Run MCP server over SSE
python -m reflex_docs_mcp.server --transport sse --host 127.0.0.1 --port 8000

Install From PyPI

pip install reflex-docs-mcp

MCP Tools

  • search_docs(query, limit?, include_content?, fuzzy?) — Full-text search with fuzzy matching

  • get_doc(slug, extract_code?) — Retrieve a full doc page by slug

  • list_pages(prefix?, limit?) — List doc pages by slug prefix

  • list_components(category?) — List all Reflex components

  • search_components(query, limit?) — Search components by name/description

  • get_component(name) — Get details about a specific component

  • get_component_props(name, filter?) — Get component props table

  • get_stats() — Database statistics

  • get_code_examples(topic, limit?) — Find Python code examples for a topic

  • decode_error(error_text, context?) — Analyze errors against docs

  • get_changelog(version?, limit?) — Fetch Reflex release notes

  • get_migration_guide(from_version, to_version) — Migration guidance between versions

  • search_api_reference(symbol) — API reference lookup

  • list_recipes(category?) — Browse recipe/tutorial pages

Local MCP Config (VS Code)

The repository includes a ready-to-use config at .vscode/mcp.json that runs the server with the local venv.

Global Install MCP Config (VS Code)

If you install the package globally with pip, use one of these in .vscode/mcp.json:

Using module invocation:

{
  "servers": {
    "reflex-docs": {
      "type": "stdio",
      "command": "python3",
      "args": ["-m", "reflex_docs_mcp.server"],
      "env": {
        "REFLEX_DOCS_AUTO_INDEX": "true"
      }
    }
  },
  "inputs": []
}

Using the CLI entry point:

{
  "servers": {
    "reflex-docs": {
      "type": "stdio",
      "command": "reflex-docs-mcp",
      "args": [],
      "env": {
        "REFLEX_DOCS_AUTO_INDEX": "true"
      }
    }
  },
  "inputs": []
}

Project Layout

├── main.py                 # MCP stdio entry point
├── src/reflex_docs_mcp/
│   ├── models.py           # Pydantic data models
│   ├── database.py         # SQLite + FTS5 operations
│   ├── parser.py           # Markdown parser
│   ├── http.py            # Shared HTTP client with TTL cache
│   ├── indexer.py          # Docs cloning & indexing
│   └── server.py           # MCP server (stdio + SSE)
├── render.yaml             # Render deployment config
├── Procfile                # Process definition
└── test.py                 # Groq + MCP demo client

Demo (Optional)

The demo client uses Groq (OpenAI-compatible API) and the MCP Python client.

pip install reflex-docs-mcp[demo]
cp env.example .env
# Add GROQ_API_KEY to .env
python test.py

Notes

  • env.example contains Groq settings.

  • The indexer writes to data/reflex_docs.db by default.

  • On startup, the server auto-builds the index if missing. Controls:

  • REFLEX_DOCS_AUTO_INDEX (default: true)

  • REFLEX_DOCS_DOCS_SRC (path to clone docs into, default: docs_src)

  • REFLEX_DOCS_SKIP_CLONE / REFLEX_DOCS_FORCE_CLONE

  • REFLEX_DOCS_KEEP_EXISTING

  • REFLEX_DOCS_CACHE_TTL (default: 300) — Seconds for search cache TTL

  • REFLEX_DOCS_HTTP_TIMEOUT (default: 10) — Seconds for outbound HTTP requests

  • REFLEX_DOCS_MAX_SEARCH_LIMIT (default: 30) — Hard cap on search result limit

  • REFLEX_DOCS_ENABLE_LIVE_FETCH (default: true) — If false, disable live HTTP fetches

License

MIT

Available Tools

14 tools
decode_errorA

Analyze a Reflex error and find relevant documentation.

Args: error_text: The error message or traceback text context: Optional additional context about what you were doing

Returns: Dict with error class, rx symbols, relevant docs, and search queries used

Inputs: error_text (required), context. Returns: object

get_changelogA

Fetch Reflex release notes from the changelog.

Args: version: Specific version to fetch (e.g., "0.6.1"). Empty for latest releases. limit: Number of recent releases to return if version is empty (default: 5)

Returns: Dict with requested version, count, source URL, and release sections

Inputs: version, limit. Returns: object

get_code_examplesA

Find Python code examples from Reflex docs related to a topic.

Args: topic: Topic to search for (e.g., "state vars", "event handlers") limit: Maximum number of code examples to return (default: 5)

Returns: Dict with topic, count, and list of code examples with source info

Inputs: topic (required), limit. Returns: object

get_componentA

Get detailed information about a specific Reflex component.

Args: name: Component name (e.g., "rx.box", "rx.button", "box", "button") The "rx." prefix is optional.

Returns: Component info with name, category, description, and documentation URL

Example: get_component("rx.box") get_component("button")

Inputs: name (required). Returns: result

get_component_propsA

Get the props table for a specific Reflex component.

Args: name: Component name (e.g., "rx.button", "button"). The "rx." prefix is optional. filter: Optional filter string to match prop names or descriptions

Returns: Dict with component name, total props count, and filtered props list

Inputs: name (required), filter. Returns: object

get_docA

Retrieve a full documentation page by its slug.

Args: slug: Document slug (e.g., "library/layout/box", "components/rendering_iterables") extract_code: If True, return only fenced code blocks from the page

Returns: Full document with title, URL, and all sections with markdown content

Example: get_doc("library/layout/box") get_doc("state/overview", extract_code=True)

Inputs: slug (required), extract_code. Returns: result

get_migration_guideA

Get migration guidance between two Reflex versions.

Args: from_version: The version you're migrating from (e.g., "0.5.0") to_version: The version you're migrating to (e.g., "0.6.0")

Returns: Dict with breaking changes, relevant docs, and changelog section

Inputs: from_version (required), to_version (required). Returns: object

get_statsA

Get statistics about the indexed documentation.

Returns: Dictionary with counts of pages, sections, and components

Inputs: no named parameters. Returns: object

list_componentsA

List all documented Reflex components.

Args: category: Optional category filter (e.g., "layout", "forms", "data-display")

Returns: List of components with name, category, description, and documentation URL

Example: list_components() list_components("layout") list_components("forms")

Inputs: category. Returns: result

list_pagesA

List documentation pages, optionally filtered by slug prefix.

Args: prefix: Optional slug prefix (e.g., "library/", "state/") limit: Maximum number of results to return (default: 200)

Returns: List of pages with slug, title, and URL

Inputs: prefix, limit. Returns: result

list_recipesA

Browse recipe and tutorial pages from Reflex docs.

Args: category: Optional category filter (e.g., "auth", "database")

Returns: Dict with category filter, count, and list of recipes with summaries

Inputs: category. Returns: object

search_api_referenceA

Look up API reference documentation for a Reflex symbol.

Args: symbol: The symbol to look up (e.g., "rx.State", "rx.button", "EventHandler")

Returns: Dict with symbol info, headings, code blocks, and tables

Inputs: symbol (required). Returns: object

search_componentsA

Search components by name or description.

Args: query: Search query (e.g., "button", "layout", "table") limit: Maximum number of results to return (default: 20)

Returns: List of matching components with name, category, description, and documentation URL

Inputs: query (required), limit. Returns: result

search_docsA

Search Reflex documentation by keyword or natural language query.

Args: query: Search query (e.g., "rx.foreach", "state management", "styling") limit: Maximum number of results to return (default: 10) include_content: If True, include full page content in each result (max 5 results) fuzzy: If True, enable prefix expansion for broader matching (default: True)

Returns: List of matching documentation sections with slug, title, score, snippet, and URL

Example: search_docs("rx.foreach") search_docs("how to style components", fuzzy=True)

Inputs: query (required), limit, include_content, fuzzy. Returns: result

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Deploy Server
    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    TDQS

    A4.2/5.0

    Scored across 14 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose: error decoding, changelog, code examples, component details, props, full doc pages, migration guides, stats, and various listing/search functions. Even overlapping tools like get_component, list_components, and search_components are complementary with clear boundaries.

    Naming Consistency5/5

    All tools follow a consistent verb_noun pattern with snake_case (e.g., decode_error, get_changelog, list_components, search_docs). No mixing of conventions, making the interface predictable.

    Tool Count5/5

    14 tools is well-scoped for a documentation server. It provides a comprehensive set of operations without being overwhelming, covering error analysis, browsing, searching, and retrieval.

    Completeness5/5

    The tool set covers all major documentation needs: error-assisted lookup, search by keyword/API symbol/component, browsing components/pages/recipes, retrieving full docs and migration guides, and obtaining statistics. No obvious gaps for typical agent interactions with a framework's docs.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Resources

    Related MCP Connectors

    Related MCP Servers