BSL Atlas MCP
A Dockerized MCP server for searching 1C dumps: structural index, call graph and semantic code search
- 79 ★ on the repository
- Last commit
- AGPL-3.0
Extended access
Why this level
- Runs a container that reads every file in the dump
- In full mode sends code fragments to an embeddings provider
Install
- Works in
- Services
Manual install
curl -O https://raw.githubusercontent.com/Arman-Kudaibergenov/bsl-atlas/master/docker-compose.yml
curl -O https://raw.githubusercontent.com/Arman-Kudaibergenov/bsl-atlas/master/.env.example
cp .env.example .env
docker compose up -dSet SOURCE_PATH and INDEXING_MODE in .env before starting.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Runs a container that reads every file in the dump
- In full mode sends code fragments to an embeddings provider
This is third-party code. Review the repository files before installing.
What it does
The server indexes the XML and BSL dump of a configuration or extension into SQLite with full-text search. The agent looks up routines by name, lists module functions with call context, and finds metadata objects and their structure. Fast mode needs no external APIs; full mode adds vector search over code and help through a chosen embeddings provider. After a new dump the index refreshes via a reindex request, and short LLM-generated descriptions can be imported for uncommented functions.
Who it is for. For 1C developers who want dump search running quickly in Docker without external keys.
Good fit when
- The agent needs to find routines and metadata objects in a dumped configuration
- You want a mode with no cloud APIs, only the structural index
- You want meaning-based function search and have an embeddings provider
Not a fit when
- No Docker, or the source folder cannot be mounted into the container
- You need a live infobase rather than a file dump
Example request
Use bsl-atlas to find the routine that fills the tabular section and show where it is called fromMCP
- Transport
- http
- Authentication
- not required
Environment variables
SOURCE_PATHrequired- Host folder with the configuration dump
INDEXING_MODErequired- Indexing mode: fast or full
Limitations
Works only with a file dump, and each new dump needs a reindex. On Windows, Docker Desktop struggles with paths containing spaces or Cyrillic, so the author suggests a symlink. Full mode needs an embeddings provider key such as OpenRouter, or a local Ollama. The repository is AGPL-3.0.
How to disable. Remove the bsl-atlas block from .mcp.json or claude_desktop_config.json and stop the container with docker compose down.
FAQ
How does fast mode differ from full?
Fast gives structural search with no external APIs. Full adds semantic search over code and help and needs an embeddings provider.
Atlas says the source folder is empty. What now?
It is almost always a wrong bind mount: check SOURCE_PATH, and on Windows replace a Cyrillic path with a symlink.
Discussion
No questions or reviews yet.
Related
All 1C:Enterprise toolsA local read-only MCP server for 1C code search on the Zoekt engine: fast lexical and regex search over your own configuration dumps
1C configuration graph MCP server
1C Configuration MCP Server
An MCP server builds a metadata and call graph from a 1C configuration's XML and BSL export, with optional semantic search via an external embedder and Qdrant
A read-only MCP service for hybrid 1C BSL code search: BM25 plus vector search, deduplication and reranking
Semantic search over 1C platform help
1c-sntx-sem
An MCP server for Cursor and Claude Desktop with hybrid embeddings and BM25 search over 1C platform help, the query language and the standard subsystems library
A SQLite code index with MCP tools for search, call graphs and metadata, including 1C dumps from Designer and EDT
An MCP server for Claude Code: full-text search over multiple 1C XML+BSL configuration exports in a local SQLite FTS5 database
README in short
The README walks through dumping the configuration to files, downloading docker-compose.yml and .env.example, setting the source path and mode, starting the container and connecting MCP over HTTP. It covers Windows mount issues and supported folder layouts separately. It lists the search tools, recommended embedding models and a way to add LLM-generated function descriptions.