BSL Analyzer
A Rust-based Language Server for BSL with code diagnostics, incremental analysis and a built-in MCP server for AI agents
Medium risk
We rate an entry medium when the tool runs code, makes network calls or reads project files. Check what exactly it does before installing.
Why this level
- Live MCP tools execute code and queries against a live 1C infobase through --onec-url
- The workspace profile reads the source code of the whole project
Install
Manual install
curl -fsSL https://raw.githubusercontent.com/itrous/bsl-analyzer/develop/scripts/install.sh | bashBinary install on Linux and macOS.
This is third-party code. Review the repository files before installing.
What it does
BSL Analyzer applies the rust-analyzer architecture to the 1C:Enterprise language: incremental computation through Salsa, parsing into a full-fidelity CST and a layered HIR with type inference. That gives semantic hover and completion aware of an expression's real type, symbol-based find references and rename instead of text search, and flow-sensitive diagnostics such as unreachable code. Besides an LSP for editors, the binary can analyze a project from the console with SARIF or JSONL output for CI, and it starts an MCP server that gives AI agents search over 1C code, platform reference, SDBL queries and debugging. Platform method and type data comes from the built-in 8.3.27 help.
Who it is for. For 1C developers who want modern BSL diagnostics and completion in their editor, and for those connecting an AI agent to platform code and reference through MCP.
Good fit when
- You need BSL code-quality diagnostics and SARIF reports for CI
- You want a modern LSP with semantic hover and completion for VS Code, Cursor or Zed
- You need an MCP server so an agent can search project code, platform reference and SDBL queries
Not a fit when
- The project uses a 1C platform older than 8.3.27: method and type facts come only from that version of the reference
- You need a stable production tool: the project is explicitly described as experimental
Example request
Check the project for unreachable branches and lost Return statements, show the report in SARIF formatLimitations
The author explicitly calls the project an experimental Language Server. Analysis and diagnostics target the full 8.3.27 API, and earlier compatibility modes are not checked. Live MCP tools need the infobase URL and credentials, and the reference profile optionally needs an EMBEDDING_URL and a 1C:Naparnik token.
How to disable. Remove the server with claude mcp remove or the equivalent for your client, uninstall the bsl-analyzer-lsp editor extension and delete the installed binary from PATH.
MCP
- Transport
- stdio
- Authentication
- API key
| Environment variables | |
|---|---|
| NAPARNIK_TOKEN secret | 1C:Naparnik token for extended reference in the reference profile. |
| EMBEDDING_URL | Embeddings service address for semantic search over the reference. |
Security check
- Live MCP tools execute code and queries against a live 1C infobase through --onec-url
- The workspace profile reads the source code of the whole project
README in short
The README explains the idea: apply the rust-analyzer architecture to BSL to get incremental analysis, semantic hover and flow-sensitive diagnostics. It covers installation via script for Linux/macOS/Windows and a separate editor extension, the lsp, analyze and mcp serve run modes, MCP auto-install through mcp install with reference and workspace profiles, and links to detailed architecture and configuration docs.
FAQ
What is the difference between the reference and workspace profiles?
reference gives platform and ITS reference without tying to a specific repository, workspace searches project code, metadata and a live 1C base through --onec-url.
Which platform versions are supported?
The project is developed and tested only against 1C:Enterprise 8.3.27, and API facts come from that version's reference.
Related
A skills library that gives coding agents a development process: brainstorming, planning, TDD, subagents and code review
Skills for real engineers by Matt Pocock
Skills For Real Engineers
Small composable skills for engineering with agents: plan grilling, TDD, bug diagnosis, code review and architecture
GitHub toolkit for spec-driven development: the specify CLI adds agent commands and skills to a project, from principles to implementation
Reference MCP servers
Model Context Protocol servers
Official reference MCP servers: Filesystem, Fetch, Git, Memory, Sequential Thinking, Time and Everything