Code intelligence, grounded in your repo

The code graph that lives next to git.

Find callers, trace dependencies, and give your AI agent context from your codebase. Travsr connects the structure of your code to the tools you already use.

npm install -g @travsr.com/travsr

Open source · Apache-2.0 · Runs locally

Follow the connections

What depends on this code?

Explore callers, dependencies, and paths through an indexed graph. Illustration with sample code.

Explore the product
LocalYour graph stays on your machine
26MCP tools in single-repo mode
14Programming language groups
Apache-2.0Open source core

The Travsr idea

01 / 04

Migratory birds cross 10,000 kilometers without a wrong turn.

Not instinct. Mapping. We built Travsr on that idea.

Start with one repository

Install. Index. Ask.

Run travsr init inside your repo to build its graph. Travsr detects installed coding tools, configures project MCP settings where supported, and prints any remaining setup.

Follow the quickstart →

Want agent guidance files too? Opt in with travsr connect --rules.

Example session with sample output. The same graph powers your editor. Run Travsr: Show Graph in VS Code to explore callers and dependencies visually. See the extension.

One graph, several ways to work

Understand the change before you make it.

Trace the impact

Find callers and dependencies, inspect a file's blast radius, and follow execution paths through indexed relationships.

Give agents focused context

Retrieve relevant symbols and source snippets within a token budget. Use MCP from your preferred coding assistant.

Keep your graph current

The running daemon watches saved files. Git hooks reconcile commits, merges, and branch checkouts. Inspect indexing status and graph health to verify progress.

Work across repositories

Register multiple repos and query them through one MCP server. Scope requests to a repo to keep answers focused.

How the graph works →

Nature, translated

A map becomes useful when every point connects.

The same structure can tell two stories. Explore how a migratory form becomes a navigable code graph.

A closer look

In your editor. Alongside your agent.

Explore these sample previews, then connect Travsr to your own code.

Follow callers and dependencies

Inspect callers with CodeLens, check blast radius before editing, and explore or copy graph context into chat from VS Code.

Illustrative preview with sample data. Explore the extension →

Bring the graph into the conversation

MCP tools give your assistant repository context and source snippets within a token budget.

Illustrative preview with sample data. Browse the MCP tools →

See what is indexed

Inspect graph statistics and manage registered repositories from the editor.

31
▸ 3 callers↳ 2 dependencies·fresh @ f8d3939
export function checkRateLimit(req: Request)
Graph stats
Graph stats

Live metrics for the indexed graph.

Nodes
2481
Edges
2609
Schema
1
DB size
2.1 MB
Last indexed
2 min ago
Refresh
Registered repos
Registered repos

3 repos in ~/.travsr/registry.json · 1 stale.

Prune stale (1) Refresh
NameDB PathStatusRemove
travsr~/learning/travsr/.travsr/graph.dbactive
travsr-web~/learning/travsr-web/.travsr/graph.dbactive
old-api~/archive/old-api/.travsr/graph.dbstale

Illustrative preview with sample data. See editor features →

Code graph preview

Sample data. Scroll horizontally to explore the full preview.

Language support

Structure first.
Deeper analysis when you need it.

Tree-sitter parses all 14 language groups below. Semantic analysis adds resolved relationships, with coverage depending on the language and analyzer.

Configuration and documentation belong here too: index JSON / JSONC, YAML, TOML, XML, and Markdown / MDX alongside your source code.

Language setup and requirements →

Built-in semantic analysis

TypeScript / JavaScript · Rust · Python

See all languages and setup
  • TypeScript / JavaScriptBuilt in
  • RustBuilt in
  • PythonBuilt in
  • GoInstall analyzer
  • JavaInstall analyzer
  • KotlinInstall analyzer
  • ScalaInstall analyzer
  • C#Install analyzer
  • C / C++Install analyzer
  • RubyInstall analyzer
  • PHPInstall analyzer
  • SwiftInstall analyzer
  • DartInstall analyzer
  • Objective-CInstall analyzer · macOS

For an additional analyzer, run travsr lang install <lang>. Objective-C semantic analysis requires macOS.

Search your way

Start with a symbol.
Or start with a question.

Structural queries work without embeddings. For broader natural-language discovery, add optional local embeddings with travsr embed init. Search can rank candidates; analyzers determine the code relationships.

Explore search and setup commands →

Bring your own codebase

Your next answer starts with your repo.

Install Travsr, connect your tools, and follow the code.