Skip to main content
Glama
bigbag

LLMDoc

by bigbag
OverviewSchemaRelated ServersScoreDiscussions
Python
Remote

LLMDoc

CI pypi downloads versions license

MCP server with RAG (BM25) for llms.txt documentation. Provides semantic search across documentation sources with automatic background refresh.

Features

  • llms.txt support - Automatically parses and indexes documentation from llms.txt files

  • Hybrid two-stage search - DuckDB FTS with Porter stemming for broad recall, BM25 reranking for precision

  • Named sources - Configure sources with names like fast_mcp:https://... for easy filtering

  • Source filtering - Search across all sources or filter by specific source name

  • Persistent storage - DuckDB-based index that survives restarts

  • Background refresh - Configurable auto-refresh interval (default: 6 hours)

  • Source attribution - Every search result includes source name and URL

Related MCP server: team-docs-mcp

Quick Start

  1. Add to Claude Code (~/.claude/claude_code_config.json):

{
  "mcpServers": {
    "llmdoc": {
      "command": "uvx",
      "args": ["llmdoc"],
      "env": {
        "LLMDOC_SOURCES": "fast_mcp:https://gofastmcp.com/llms.txt"
      }
    }
  }
}

  1. Restart Claude Code - the server will automatically fetch and index documentation.

  2. Ask Claude questions like "How do I create a tool in FastMCP?" and it will search the indexed docs.

What is llms.txt?

llms.txt is a specification for providing LLM-friendly documentation. Websites add a /llms.txt markdown file to their root directory containing curated, concise content optimized for AI consumption. LLMDoc indexes these files and their linked documents to enable semantic search.

Example sources:

Installation

# Run directly with uvx (no install needed)
uvx llmdoc

# Or install with uv
uv tool install llmdoc

# Or install with pip
pip install llmdoc

# Or install with pipx
pipx install llmdoc

Configuration

Source Format

Sources can be specified in two formats:

  • Named: name:url - e.g., fast_mcp:https://gofastmcp.com/llms.txt

  • Unnamed: Just the URL - name is auto-generated from domain

Named sources allow you to filter search results by source name.

Environment Variables

# Comma-separated list of sources (named or unnamed)
export LLMDOC_SOURCES="fast_mcp:https://gofastmcp.com/llms.txt,pydantic_ai:https://ai.pydantic.dev/llms.txt"

# Optional: Custom database path (default: ~/.llmdoc/index.db)
export LLMDOC_DB_PATH="/path/to/index.db"

# Optional: Refresh interval in hours (default: 6)
export LLMDOC_REFRESH_INTERVAL="6"

# Optional: Max concurrent document fetches (default: 5)
export LLMDOC_MAX_CONCURRENT="5"

# Optional: Skip refresh on startup (default: false)
export LLMDOC_SKIP_STARTUP_REFRESH="true"

# Optional: Disable FTS indexing and use pure BM25 (default: true)
export LLMDOC_ENABLE_FTS="false"

Config File

Create llmdoc.json in the working directory:

{
  "sources": [
    "fast_mcp:https://gofastmcp.com/llms.txt",
    "pydantic_ai:https://ai.pydantic.dev/llms.txt"
  ],
  "db_path": "~/.llmdoc/index.db",
  "refresh_interval_hours": 6,
  "max_concurrent_fetches": 5,
  "skip_startup_refresh": false,
  "enable_fts": true
}

Or with explicit name/url objects:

{
  "sources": [
    {"name": "fast_mcp", "url": "https://gofastmcp.com/llms.txt"},
    {"name": "pydantic_ai", "url": "https://ai.pydantic.dev/llms.txt"}
  ]
}

Running the Server

LLMDoc uses stdio transport and is designed to be launched by MCP clients. Configure it in your MCP client (see below), and the client will start the server automatically.

For manual testing:

# Using uvx
uvx llmdoc

# Or as module
python -m llmdoc

MCP Tools

  • search_docs(query, limit, source) - Search documentation and return relevant passages with source URLs. Optional source parameter filters by source name (e.g., fast_mcp)

  • get_doc(url, offset, limit) - Get document content with pagination support for large documents. Parameters: offset (default: 0) start position in bytes, limit (default: 50000, max: 100000) max bytes per call. Returns pagination metadata (has_more, total_length)

  • get_doc_excerpt(url, query, max_chunks, context_chars) - Get relevant excerpts from a large document matching a query

  • list_sources() - List all configured documentation sources with statistics

  • refresh_sources() - Manually trigger a refresh of all documentation

MCP Resources

  • doc://sources - Returns JSON with configured sources list and refresh interval

Adding to MCP Clients

Claude Code

Add to ~/.claude/claude_code_config.json:

{
  "mcpServers": {
    "llmdoc": {
      "command": "uvx",
      "args": ["llmdoc"],
      "env": {
        "LLMDOC_SOURCES": "fast_mcp:https://gofastmcp.com/llms.txt,pydantic_ai:https://ai.pydantic.dev/llms.txt"
      }
    }
  }
}

Standard MCP Configuration

Add to your MCP client's configuration file:

{
  "mcpServers": {
    "llmdoc": {
      "command": "uvx",
      "args": ["llmdoc"],
      "env": {
        "LLMDOC_SOURCES": "fast_mcp:https://gofastmcp.com/llms.txt"
      }
    }
  }
}

Example Usage

Once configured, the LLM can use these tools:

User: How do I create a tool in FastMCP?

LLM: [calls search_docs("create tool FastMCP")]

Result:
[
  {
    "title": "Tools",
    "snippet": "Creating a tool is as simple as decorating a Python function with @mcp.tool...",
    "url": "https://gofastmcp.com/servers/tools.md",
    "source": "fast_mcp",
    "source_url": "https://gofastmcp.com/llms.txt",
    "score": 12.5
  }
]

Filtering by Source

You can filter results to a specific documentation source:

User: How do I create an agent in PydanticAI?

LLM: [calls search_docs("create agent", source="pydantic_ai")]

Result:
[
  {
    "title": "Agents",
    "snippet": "Agents are the primary interface for interacting with LLMs in PydanticAI...",
    "url": "https://ai.pydantic.dev/agents.md",
    "source": "pydantic_ai",
    "source_url": "https://ai.pydantic.dev/llms.txt",
    "score": 10.2
  }
]

Getting Full Document Content

Use get_doc to retrieve document content (supports pagination for large documents):

LLM: [calls get_doc("https://ai.pydantic.dev/agents.md")]

Result:
{
  "title": "Agents",
  "content": "# Agents\n\nAgents are the primary interface for interacting with LLMs in PydanticAI...",
  "url": "https://ai.pydantic.dev/agents.md",
  "source": "pydantic_ai",
  "source_url": "https://ai.pydantic.dev/llms.txt",
  "offset": 0,
  "length": 5432,
  "total_length": 5432,
  "has_more": false
}

Architecture

+------------------+
|    MCP Client    |
| (Claude, Cursor) |
+--------+---------+
         | stdio
         v
+------------------+     +------------------+     +------------------+
|  FastMCP Server  |---->|  Document Store  |<----|Document Fetcher  |
|                  |     |    (DuckDB)      |     | (async HTTP)     |
|  - search_docs   |     |                  |     |                  |
|  - get_doc       |     |  - Persistence   |     | - llms.txt parse |
|  - list_sources  |     |  - Deduplication |     | - HTML→Markdown  |
|  - refresh       |     |  - Change detect |     | - Concurrent     |
+--------+---------+     +------------------+     +------------------+
         |
         v
+------------------+
|   BM25 Index     |
|   (in-memory)    |
|                  |
|  - Chunking      |
|  - Tokenization  |
|  - Scoring       |
+------------------+

LLMDoc fetches documentation from llms.txt sources, stores it in DuckDB, and provides fast BM25 search through the MCP protocol.

How It Works

Document Fetching

When configured with documentation sources, LLMDoc:

  1. Parses llms.txt files to discover all linked documents

  2. Fetches each document concurrently (with rate limiting)

  3. Converts HTML pages to Markdown automatically

  4. Extracts titles from the first H1 heading

Indexing

Documents are processed for efficient search:

  1. Chunking: Large documents are split into ~500 character chunks at sentence boundaries

  2. Tokenization: Text is lowercased and stopwords are removed

  3. Indexing: BM25 algorithm indexes all chunks for relevance scoring

Search

LLMDoc uses a hybrid two-stage retrieval approach:

Stage 1 - DuckDB FTS (Recall):

  1. Query is processed by DuckDB's full-text search with Porter stemming

  2. "running" matches "run", "café" matches "cafe"

  3. Top 100 candidate chunks are retrieved

Stage 2 - Python BM25 (Precision):

  1. Candidates are re-scored using exact-match BM25

  2. Documents with exact query terms rank higher

  3. Results are deduplicated by document URL

  4. Top results are returned with relevance scores and snippets

Background Refresh

LLMDoc automatically keeps documentation up-to-date:

  • Checks for staleness on startup

  • Refreshes every 6 hours (configurable)

  • Uses content hashing to skip unchanged documents

  • Removes documents no longer in llms.txt

Technical Details

Hybrid Two-Stage Search

LLMDoc combines DuckDB's native FTS with Python BM25 for optimal search quality:

Stage 1 - DuckDB FTS:

  • Porter stemming normalizes words (running → run, documents → document)

  • Accent handling (café → cafe)

  • 571 built-in English stopwords

  • Fast candidate retrieval using native C implementation

Stage 2 - Python BM25:

  • BM25Okapi algorithm from rank_bm25 library

  • Exact term matching boosts precise matches

  • Term frequency saturation, document length normalization, IDF weighting

  • Thread-safe using threading.RLock()

Chunking Strategy

Documents are chunked using a multi-level approach:

  1. Paragraph splitting: First split on double newlines (\n\n)

  2. Sentence-boundary aware: Long paragraphs split at .!? followed by whitespace

  3. Overlap: 100 character overlap between chunks maintains context

Configuration:

  • chunk_size: 500 characters (default)

  • chunk_overlap: 100 characters (default)

Database Schema

DuckDB stores documents and chunks:

CREATE TABLE documents (
    id INTEGER PRIMARY KEY,
    source_name TEXT NOT NULL,    -- e.g., 'fast_mcp'
    source_url TEXT NOT NULL,     -- llms.txt URL
    doc_url TEXT NOT NULL UNIQUE, -- document URL
    title TEXT,
    content TEXT NOT NULL,
    content_hash TEXT NOT NULL,   -- SHA256 for change detection
    updated_at TIMESTAMP NOT NULL
)

CREATE TABLE chunks (
    id INTEGER PRIMARY KEY,
    doc_id INTEGER NOT NULL,      -- references documents.id
    content TEXT NOT NULL,        -- chunk text for FTS indexing
    start_pos INTEGER NOT NULL,   -- position in original document
    end_pos INTEGER NOT NULL
)

FTS index on chunks table with Porter stemmer for hybrid search.

Concurrency Model

LLMDoc supports multiple concurrent instances:

  • Read operations: Multiple instances can search simultaneously (read-only DuckDB mode)

  • Write operations: Single instance holds exclusive lock during refresh

  • Graceful handling: If refresh is locked, operation skips with status message

Document fetching uses asyncio.Semaphore to limit concurrent HTTP requests (default: 5).

Stopwords

Two stopword lists are used:

  • DuckDB FTS (Stage 1): 571 built-in English stopwords

  • Python BM25 (Stage 2): 209 custom stopwords including articles, prepositions, pronouns, auxiliaries, and common verbs

License

MIT License - see LICENSE file.

Available Tools

5 tools
get_docGet DocumentA
Read-onlyIdempotent

Get document content with pagination support for large documents.

For documents larger than 50KB, use offset/limit to paginate through content. The response includes has_more=True if more content is available. For targeted retrieval, use get_doc_excerpt instead.

Args: url: The URL of the document (as returned by search_docs). offset: Start position in bytes (default: 0). limit: Max bytes to return per call (default: 50000, max: 100000).

Returns: Document with content chunk, pagination metadata (offset, length, total_length, has_more).

ParametersJSON Schema
NameRequiredDescriptionDefault
urlYesThe URL of the document (as returned by search_docs)
limitNoMax bytes to return (default 50000, max 100000)
offsetNoStart position in bytes (for pagination)

Output Schema

ParametersJSON Schema
NameRequiredDescription
urlYes
titleYes
lengthNo
offsetNo
sourceYes
contentYes
has_moreNo
source_urlYes
total_lengthNo

TDQS

A4.7/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint=true and idempotentHint=true, establishing the safe read-only nature. The description adds valuable behavioral context: pagination mechanics, the 50KB threshold, response includes has_more, and metadata fields (offset, length, total_length, has_more). It does not describe error behavior or rate limits, but the annotations cover the primary safety aspects.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is well-structured: opening sentence states purpose, followed by concise usage guidance, a clear Args section, and a Returns section. Every sentence contributes necessary information without redundancy, and key details are front-loaded.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The tool has moderate complexity with pagination. The description covers pagination triggers, parameter semantics, response format, and an alternative tool. It leaves no obvious gaps for a read-only fetch operation, especially given the existing annotations and output schema.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so baseline is 3. The description adds semantics beyond individual parameter descriptions by explaining when to use offset/limit (documents >50KB) and what has_more indicates in the response. This contextual usage guidance enriches the parameter meaning.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's function: 'Get document content' with pagination support. It distinguishes from the sibling tool get_doc_excerpt by noting 'For targeted retrieval, use get_doc_excerpt instead.'

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides explicit when-to-use guidance: 'For documents larger than 50KB, use offset/limit to paginate' and explicitly names the alternative for targeted retrieval. This gives clear context for choosing between tools.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

get_doc_excerptGet Document ExcerptA
Read-onlyIdempotent

Get relevant excerpts from a large document matching a query.

Use this instead of get_doc for large documents. Returns targeted excerpts based on BM25 relevance to your query.

Args: url: The URL of the document. query: Query to find relevant sections within the document. max_chunks: Maximum number of chunks to return (default: 5). context_chars: Extra context characters around each chunk (default: 500).

Returns: Document metadata with list of relevant excerpts, each containing content, position, and relevance score.

ParametersJSON Schema
NameRequiredDescriptionDefault
urlYesThe URL of the document
queryYesQuery to find relevant sections within the document
max_chunksNoMaximum chunks to return
context_charsNoExtra context chars around each chunk

Output Schema

ParametersJSON Schema
NameRequiredDescription
urlYes
titleYes
sourceYes
excerptsYes
source_urlYes
total_lengthYes

TDQS

A4.4/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already indicate readOnlyHint and idempotentHint, so the safety profile is covered. The description adds behavioral context beyond annotations by mentioning the BM25 relevance algorithm and the return structure (list of excerpts with content, position, and relevance score). It does not discuss edge cases like no relevant sections, but the added context is meaningful.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is well-structured: purpose, usage, args, and returns are clearly separated. The purpose and usage sentences are front-loaded and efficient. However, the Args and Returns sections largely duplicate schema information, making the description slightly longer than necessary for an AI agent that already has structured data.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the output schema exists, the description doesn't need to detail return values, but it does anyway. It provides purpose, usage guidance, parameter defaults, and a clear return summary. The tool is simple and read-only, and the description is complete enough for an agent to select and invoke it correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100% and includes descriptions and defaults for all four parameters. The description's Args section repeats the schema information without adding new semantic meaning, so it does not compensate beyond the baseline for high schema coverage.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb ('Get'), resource ('relevant excerpts from a large document'), and scope ('matching a query'). It also clearly distinguishes itself from get_doc by explicitly recommending 'Use this instead of get_doc for large documents'.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives explicit when-to-use guidance: 'Use this instead of get_doc for large documents.' This names an alternative tool and the condition for using it. It implies when not to use it (when the document is not large, use get_doc).

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

list_sourcesList SourcesA
Read-onlyIdempotent

List all configured documentation sources with their statistics.

Use this to discover what documentation sources are available for searching. Each source has a name that can be used to filter search_docs results.

Returns: List of sources with name, url, doc_count, and last_updated.

ParametersJSON Schema
NameRequiredDescriptionDefault

No parameters

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A4.5/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint and idempotentHint, so the description need not restate safety. It adds value by specifying the return fields (name, url, doc_count, last_updated) and the relationship to search_docs.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Four short sentences cover purpose, use case, sibling relationship, and return format without redundancy. Front-loaded with the core action.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a simple read-only list operation, the description is complete: it states what is listed, how to use it, what it returns, and how it relates to search_docs. Output schema exists, so no need to detail every field.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

There are zero parameters and the empty input schema fully covers them, meeting the baseline of 4. The description's mention that sources can be used to filter search_docs indirectly clarifies parameter context.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with 'List all configured documentation sources with their statistics,' a specific verb and resource. It clearly distinguishes from siblings by focusing on enumeration and discovery, and explicitly links source names to search_docs filtering.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

'Use this to discover what documentation sources are available for searching' provides direct usage context. It also notes the names can filter search_docs results, but does not explicitly contrast with refresh_sources or other alternatives.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

refresh_sourcesRefresh SourcesA

Manually trigger a refresh of all documentation sources.

This fetches documentation from all configured llms.txt URLs and updates the local index. Use this when you need the latest documentation content.

Returns: Dictionary with refreshed_count, indexed_documents, indexed_chunks, sources, and any errors.

ParametersJSON Schema
NameRequiredDescriptionDefault

No parameters

Output Schema

ParametersJSON Schema
NameRequiredDescription
errorsNo
reasonNo
skippedNo
sourcesYes
indexed_chunksYes
refreshed_countYes
indexed_documentsYes

TDQS

A4.5/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description discloses that it fetches from 'all configured llms.txt URLs' and 'updates the local index,' which adds behavioral detail beyond the annotations (readOnlyHint=false, idempotentHint=false, destructiveHint=false). It also describes the return dictionary including counts and errors. No contradiction with annotations.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is three concise sentences with the main verb phrase front-loaded. It efficiently conveys purpose, mechanism, and return value without redundant filler. Every sentence earns its place.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a no-parameter trigger tool with an output schema, the description is complete. It covers the purpose, the scope ('all documentation sources'), the mechanism ('fetches from llms.txt URLs'), the side effect ('updates the local index'), and the return structure. No significant gaps remain.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

There are zero parameters, and the schema coverage is 100% (empty properties). The description correctly omits any parameter details. The baseline for 0 params is 4, and the description adds no unnecessary param info, making it appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a specific verb+resource: 'Manually trigger a refresh of all documentation sources.' It distinguishes from siblings (search_docs, get_doc, get_doc_excerpt, list_sources) by clearly indicating a mutating refresh operation rather than a read or query operation.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description states 'Use this when you need the latest documentation content,' providing clear context for when to invoke the tool. It does not explicitly list exclusions or alternatives, but the sibling tools are clearly different in purpose, and the usage guidance is sufficient for straightforward selection.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

search_docsSearch DocumentationA
Read-onlyIdempotent

Search documentation and return relevant passages with source URLs.

Use this tool when you need to find information about specific topics, APIs, or concepts. The search uses BM25 ranking for relevance.

Args: query: The search query to find relevant documentation. limit: Maximum number of results to return (default: 5). source: Optional source name to filter results (e.g., 'fast_mcp', 'pydantic_ai').

Returns: List of search results with title, snippet, url, source (name), source_url, and score.

ParametersJSON Schema
NameRequiredDescriptionDefault
limitNoMaximum number of results to return
queryYesThe search query to find relevant documentation
sourceNoOptional source name to filter results (e.g., 'fast_mcp', 'pydantic_ai')

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A4.1/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint=true and idempotentHint=true, so the safety profile is known. The description adds behavioral details such as BM25 ranking for relevance and the return structure (list of results with fields), going beyond what annotations provide. No contradictions.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is well-structured with a clear opening statement, usage guidance, and a formatted Args/Returns section. It is reasonably concise, though the Args section duplicates schema descriptions to some extent, making it slightly longer than necessary.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The description covers purpose, usage context, ranking algorithm, parameter details, and return fields, making it sufficiently complete for a search tool. It could mention edge cases (e.g., behavior when no results) but given the output schema exists and annotations are strong, it is adequately complete.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, and the description essentially repeats the parameter details (query, limit, source) without adding new meaning. Baseline for high coverage is 3; description does not enhance understanding beyond the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool searches documentation and returns relevant passages with source URLs, distinguishing it from siblings like get_doc and get_doc_excerpt by focusing on general search rather than retrieving specific content. It uses a specific verb ('search') and resource ('documentation').

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Provides explicit guidance on when to use the tool ('when you need to find information about specific topics, APIs, or concepts'), but does not mention exclusions or alternatives. Clear context but no explicit when-not or alternative comparisons.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.

  1. 5 tool updatesv0.3.1
    • First observedget_doc
    • First observedget_doc_excerpt
    • First observedlist_sources
    • First observedrefresh_sources
    • First observedsearch_docs
Install Server
RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

A4.4/5.0
Disambiguation5/5

Each tool has a clearly distinct purpose: search_docs for global search, get_doc for full content with pagination, get_doc_excerpt for targeted extraction, list_sources for source discovery, and refresh_sources for index updates. The overlap between search_docs and get_doc_excerpt is adequately clarified by their descriptions.

Naming Consistency5/5

All tool names follow a consistent verb_noun snake_case pattern (search_, get_, list_, refresh_), making the API predictable and easy to navigate.

Tool Count5/5

With 5 tools, the server is well-scoped for a documentation search and retrieval service. Each tool is necessary and covers a distinct operation without redundancy.

Completeness4/5

The core workflow is covered: discover sources, search, retrieve full or excerpted content, and refresh the index. Minor gaps exist around source management (e.g., adding/removing sources) and listing all documents within a source, but these are not critical for standard usage.

Maintenance

ActivityInactive
ResponsivenessNo issues

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

Related MCP Servers

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/bigbag/llmdoc'

If you have feedback or need assistance with the MCP directory API, please join our Discord server