Skip to main content

SystemVerilog Navigation MCP service for AI agents

Project description

sv-nav

SystemVerilog navigation MCP service for AI agents.

Wraps slang-server LSP capabilities into an MCP (Model Context Protocol) service, providing semantic APIs — AI agents can query by signal name and hierarchy path instead of file URI + line/column, and get structured, deduplicated results without dealing with LSP protocol details.

Installation

pip install sv-nav-mcp

Requirements

Configuration

Add to your project's .claude/settings.json or global ~/.claude/settings.json:

{
  "mcpServers": {
    "sv-nav": {
      "command": "sv-nav-mcp"
    }
  }
}

With verbose logging (output to stderr):

{
  "mcpServers": {
    "sv-nav": {
      "command": "sv-nav-mcp",
      "args": ["--verbose"]
    }
  }
}

Tools

Lifecycle

Tool Description
sv_init Initialize with top_module and build_file (.f). Required first call.
sv_refresh Refresh file content after external changes. Optional file to refresh a single file.
sv_reinit Full server restart. Expensive, use sparingly.

Navigation

Tool Description
sv_search Search symbols by regex. Global or scoped to a module/hierarchy path.
sv_definition Get symbol definition with hover content. For modules/instances, also returns ports and params.
sv_refs Find all references to a symbol or module.
sv_hierarchy Browse instance hierarchy tree (Build mode).

Tracing

Tool Description
sv_cone Trace signal driver/load cone, automatically merging duplicate nodes (Build mode).

Diagnostics

Tool Description
sv_diagnostics Get compilation errors and warnings. Returns summary grouped by diagnostic code, supports filtering by code and severity.

sv_definition: Elaborated vs Declaration Data

The data level depends on the context parameter:

  • Hierarchy path context (e.g. "top.cpu"): Returns elaborated data — instance-specific parameter values, resolved types (e.g. logic[7:0] instead of logic[W-1:0]). elaborated=True.
  • Module name context (e.g. "cpu"): Returns declaration data — source-level defaults, no type/value enrichment. elaborated=False.
  • No context: Global module search, declaration data.

Build File Format (.f)

sv_init requires a build file listing source files:

// One source file path per line
// Lines starting with // are comments
// Environment variables ($VAR, ${VAR}) are expanded
// +incdir+ directives and lines with unresolvable vars are skipped
rtl/module_a.sv
rtl/module_b.sv
$SHARED_DIR/util.sv
  • Relative paths resolve from the .f file's directory
  • No dependency ordering required

Typical Workflow

  1. sv_init(top_module="ibex_core", build_file="rtl/ibex_core.f")
  2. sv_search to find symbols of interest
  3. sv_definition to get definition details and internal structure
  4. sv_refs to find all usages
  5. sv_cone to trace signal drivers/loads
  6. sv_diagnostics to check for errors

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

sv_nav_mcp-0.1.1.tar.gz (38.9 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

sv_nav_mcp-0.1.1-py3-none-any.whl (35.6 kB view details)

Uploaded Python 3

File details

Details for the file sv_nav_mcp-0.1.1.tar.gz.

File metadata

  • Download URL: sv_nav_mcp-0.1.1.tar.gz
  • Upload date:
  • Size: 38.9 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.11.15

File hashes

Hashes for sv_nav_mcp-0.1.1.tar.gz
Algorithm Hash digest
SHA256 6386889a4d3dcb7f98707c2fe57787b2de2e849bd0e15e251af9c6c4203cbb4f
MD5 5fbe61157a8dca15258a37bac5eed228
BLAKE2b-256 bd3caae68d80771b971935b6f6a0ba5ed23012d9882ac3512f6318f43274c1bb

See more details on using hashes here.

File details

Details for the file sv_nav_mcp-0.1.1-py3-none-any.whl.

File metadata

  • Download URL: sv_nav_mcp-0.1.1-py3-none-any.whl
  • Upload date:
  • Size: 35.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.11.15

File hashes

Hashes for sv_nav_mcp-0.1.1-py3-none-any.whl
Algorithm Hash digest
SHA256 f70b27de082a708770242832abcbea6d9806dd5915cc5c84245bd504238eda57
MD5 5da068155d9a78469fcb5a6502284b70
BLAKE2b-256 cd16d233e96a3c9259d535ba27ed401067ffd8a164c3e4379985e19607fb469d

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page