Skip to main content
Glama
Sign In
Marten Docs MCP Server

Marten Docs MCP Server

by stijnVanHorenbeek
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Local

Marten Docs MCP Server

Local MCP server for focused Marten documentation retrieval.

It caches https://martendb.io/llms-full.txt, builds a local index, and exposes a narrow retrieval surface so agents search first, then progressively narrow to specific sections.

Quick install

macOS / Linux

# Install latest release
curl -fsSL https://raw.githubusercontent.com/stijnVanHorenbeek/marten_mcp/master/scripts/quickinstall.sh | sh

# Print a Copilot-compatible config snippet
curl -fsSL https://raw.githubusercontent.com/stijnVanHorenbeek/marten_mcp/master/scripts/quickinstall.sh | sh -s -- --client copilot

Windows PowerShell

# Install latest release
& ([scriptblock]::Create((irm https://raw.githubusercontent.com/stijnVanHorenbeek/marten_mcp/master/scripts/quickinstall.ps1)))

# Print a Copilot-compatible config snippet
& ([scriptblock]::Create((irm https://raw.githubusercontent.com/stijnVanHorenbeek/marten_mcp/master/scripts/quickinstall.ps1))) -Client copilot

Related MCP server: Atlas

Example SKILL.md

Example SKILL.md is included

MCP config examples

OpenCode

{
  "mcp": {
    "marten-docs": {
      "type": "local",
      "command": ["marten-docs-mcp"],
      "environment": {
        "MARTEN_MCP_CACHE_DIR": "~/.cache/marten-docs-mcp",
        "MARTEN_MCP_STORAGE_MODE": "auto",
        "MARTEN_MCP_SQLITE_PATH": "~/.cache/marten-docs-mcp/cache.db"
      }
    }
  }
}

GitHub Copilot Chat

{
  "mcpServers": {
    "marten-docs": {
      "type": "local",
      "command": "marten-docs-mcp",
      "args": [],
      "env": {
        "MARTEN_MCP_CACHE_DIR": "~/.cache/marten-docs-mcp",
        "MARTEN_MCP_STORAGE_MODE": "auto",
        "MARTEN_MCP_SQLITE_PATH": "~/.cache/marten-docs-mcp/cache.db"
      },
      "tools": ["*"]
    }
  }
}

MCP tools

  • search_docs(query, limit?, offset?)

  • search_within_page(path, query, limit?, offset?)

  • list_headings(path)

  • read_section(id, segmentIndex?, offset?, maxChars?) (returns one local segment plus compact neighbors.before/after refs)

  • read_context(id, before?, after?) (nearby refs only)

  • list_pages(prefix?, limit?) (discovery only when search is insufficient)

  • get_status()

  • refresh_docs(force?)

Recommended retrieval flow

  1. search_docs(...)

  2. Narrow with list_headings(...) and/or search_within_page(...)

  3. Read one chunk with read_section(...)

  4. Use read_context(...) only for nearby references

Broad page dumps are intentionally unsupported.

1. search_docs(query="aggregate projections", limit=3)
2. search_within_page(path="/events/projections/aggregate-projections.md", query="lifecycle", limit=3)
3. read_section(id="<id>", offset=0, maxChars=1200)
4. read_context(id="<id>", before=1, after=1)
5. repeat read_section(..., offset=<nextOffset>) while hasMore=true

Companion docs

Available Tools

8 tools
get_statusB
Read-onlyIdempotent

Get cache and index status.

Inputs: no named parameters. Returns: index, hasCache, cachePath, freshness, +2 more

list_headingsC
Read-onlyIdempotent

Inspect sections within a known page before reading a chunk.

Inputs: path (required). Returns: path, count, headings

list_pagesC
Read-onlyIdempotent

Discovery only when search does not identify a relevant page.

Inputs: limit, prefix. Returns: count, limit, pages, prefix

read_contextA
Read-onlyIdempotent

Expand nearby refs around one chunk when immediate neighbor refs are not enough.

Inputs: id (required), after, before. Returns: id, after, count, before, +1 more

read_sectionA
Read-onlyIdempotent

Read one chunk with pagination. Prefer neighbors.before/after for the next local read before running another global search.

Inputs: id (required), offset, maxChars, segmentIndex. Returns: id, path, error, kinds, +8 more

refresh_docsC

Refresh source docs and rebuild the index.

Inputs: force. Returns: force, status, metadata, refreshed, +1 more

search_docsA
Read-onlyIdempotent

Start here: global docs search. Once a relevant page is found, prefer local navigation over another global search.

Inputs: query (required), limit, offset. Returns: count, limit, query, offset, +1 more

search_within_pageA
Read-onlyIdempotent

Narrow within a known page after global search. Prefer this over another global search.

Inputs: path (required), query (required), limit, offset. Returns: path, count, limit, query, +2 more

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

    A3.6/5.0

    Scored across 8 tools

    Disambiguation5/5

    Each tool has a distinct purpose that is clearly described. Tools like search_docs, search_within_page, and read_section guide the agent through a clear search-to-read workflow without ambiguity.

    Naming Consistency5/5

    All tool names follow a consistent verb_noun pattern: list_headings, search_docs, search_within_page, read_section, read_context, list_pages, get_status, refresh_docs. The naming is predictable and uniform.

    Tool Count5/5

    With 8 tools, the count is well-scoped for a documentation server. Each tool supports a specific step in the exploration process, from search to reading to cache management.

    Completeness4/5

    The tool surface covers the full workflow: global search, local search, reading with pagination, context expansion, page listing, and cache status/refresh. A minor gap is the lack of a tool to directly fetch a page by ID or URL, but the workflow is otherwise complete.

    Maintenance

    ActivityNo data
    ResponsivenessOutdated data

    Resources

    Related MCP Connectors

    Related MCP Servers