osgrep

Natural-language search that works like grep. Fast, local, and built for coding agents.

• Semantic: Finds concepts ("where do transactions get created?"), not just strings.
• Call Graph Tracing: Map dependencies with trace to see who calls what.
• Role Detection: Distinguishes ORCHESTRATION (high-level logic) from DEFINITION (types/classes).
• Local & Private: 100% local embeddings via onnxruntime-node.
• Auto-Isolated: Each repository gets its own index automatically.
• Agent-Ready: Native output with symbols, roles, and call graphs.

Quick Start

1. Install

   npm install -g osgrep
   

2. Setup (Recommended)

   osgrep setup
   

   Downloads embedding models (~150MB) upfront. If you skip this, models download automatically on first use.

3. Search

   cd my-repo
   osgrep "where do we handle authentication?"
   

   Your first search will automatically index the repository. Each repository is automatically isolated with its own index. Switching between repos "just works" — no manual configuration needed. If the background server is running (osgrep serve), search goes through the hot daemon; otherwise it falls back to on-demand indexing.

4. Trace (Call Graph)

   osgrep trace "function_name"
   

See who calls a function (upstream dependencies) and what it calls (downstream dependencies). Perfect for impact analysis and understanding code flow.

To find the symbols in your code base: bash osgrep symbols

In our public benchmarks, osgrep can save about 20% of your LLM tokens and deliver a 30% speedup.

Claude Code Plugin

1. Run osgrep install-claude-code
2. Open Claude Code (claude) and ask it questions about your codebase.
3. Highly recommend indexing your code base before using the plugin.
4. The plugin's hooks auto-start osgrep serve in the background and shut it down on session end. Claude will use osgrep for semantic searches automatically but can be encouraged to do so.

Opencode Plugin

1. Run osgrep install-opencode
2. Open OC (opencode) and ask it questions about your codebase.
3. Highly recommend indexing your code base before using the plugin.
4. The plugin's hooks auto-start osgrep serve in the background and shut it down on session end. OC will use osgrep for semantic searches automatically but can be encouraged to do so.

Commands

osgrep search

The default command. Searches the current directory using semantic meaning.

osgrep "how is the database connection pooled?"

Options:

Flag	Description	Default
-m <n>	Max total results to return.	25
--per-file <n>	Max matches to show per file.	1
-c, --content	Show full chunk content instead of snippets.	false
--scores	Show relevance scores (0-1) for each result.	false
--min-score <n>	Filter out results below this score threshold.	0
--compact	Show file paths only (like grep -l).	false
-s, --sync	Force re-index changed files before searching.	false
-r, --reset	Reset the index and re-index from scratch.	false
Examples:

# General concept search
osgrep "API rate limiting logic"

# Deep dive (show more matches per file)
osgrep "error handling" --per-file 5

# Just give me the files
osgrep "user validation" --compact

# Show relevance scores and filter low-confidence matches
osgrep "authentication" --scores --min-score 0.5

osgrep index

Manually indexes the repository. Useful if you want to pre-warm the cache or if you've made massive changes outside of the editor.

• Respects .gitignore and .osgrepignore (see Configuration section).
• Smart Indexing: Only embeds code and config files. Skips binaries, lockfiles, and minified assets.
• Bounded Concurrency: Uses a fixed thread pool to keep your system responsive.
• Semantic Chunking: Uses TreeSitter grammars for supported languages (TypeScript, JavaScript, Python, Go, Rust, C/C++, Java, C#, Ruby, PHP, JSON, YAML, Kotlin, Swift, Dart).

Options:

Flag	Description	Default
-d, --dry-run	See what would be indexed without making changes.	false
-p, --path <dir>	Path to index (defaults to current directory).	.
-r, --reset	Remove existing index and re-index from scratch.	false
-v, --verbose	Show detailed progress with file names.	false
--allow-fallback	Also index files without TreeSitter grammar support.	false

Examples:

osgrep index                    # Index current dir
osgrep index --dry-run          # See what would be indexed
osgrep index --verbose          # Watch detailed progress (useful for debugging)
osgrep index --allow-fallback   # Include files without grammar support
osgrep index --reset            # Full re-index from scratch

osgrep serve

Runs a lightweight HTTP server with live file watching so searches stay hot in RAM.

• Keeps LanceDB and the embedding worker resident for <50ms responses.
• Watches the repo (via chokidar) and incrementally re-indexes on change.
• Health endpoint: GET /health
• Search endpoint: POST /search with { query, limit, path, rerank }
• Writes lock: .osgrep/server.json with port/pid

Options:

Flag	Description
-p, --port <port>	Port to listen on (auto-increments if not specified)
-b, --background	Run server in background and exit immediately

Port Selection (priority order):

1. Explicit -p <port> flag
2. OSGREP_PORT environment variable
3. Auto-increment from registry (last used port + 1, or 4444 if no servers)

Usage:

osgrep serve                    # Foreground, port 4444 (or next available)
osgrep serve --background       # Background mode, auto port
osgrep serve -b -p 5000         # Background on specific port

Subcommands:

osgrep serve status             # Show server status for current directory
osgrep serve stop               # Stop server in current directory
osgrep serve stop --all         # Stop all running osgrep servers

Example workflow:

# Start servers in multiple projects
cd ~/project-a && osgrep serve -b    # Starts on port 4444
cd ~/project-b && osgrep serve -b    # Starts on port 4445 (auto-increment)

# Check status
osgrep serve status

# Stop all when done
osgrep serve stop --all

Claude Code hooks start/stop this automatically; you rarely need to run it manually.

osgrep list

Lists all indexed repositories (stores) and their metadata.

osgrep list

Shows store names, sizes, and last modified times. Useful for seeing what's indexed and cleaning up old stores.

osgrep skeleton

Generates a compressed "skeleton" of a file, showing only signatures, types, and class structures while eliding function bodies.

osgrep skeleton src/lib/auth.ts

Output:

class AuthService {
  validate(token: string): boolean {
    // → jwt.verify, checkScope, .. | C:5 | ORCH
  }
}

Modes:

• osgrep skeleton <file>: Skeletonize specific file.
• osgrep skeleton <Symbol>: Find symbol in index and skeletonize its file.
• osgrep skeleton "query": Search for query and skeletonize top matches.

Supported Languages: TypeScript, JavaScript, Python, Go, Rust, Java, C#, C++, C, Ruby, PHP.

osgrep doctor

Checks installation health, model paths, and database integrity.

osgrep doctor

Performance & Architecture

osgrep is designed to be a "good citizen" on your machine:

1. Bounded Concurrency: Chunking/embedding stay within small thread pools (1–4) and capped batch sizes to keep laptops responsive.
2. Smart Chunking: Uses tree-sitter to split code by function/class boundaries, ensuring embeddings capture complete logical blocks.
3. Deduplication: Identical code blocks (boilerplate, license headers) are embedded once and cached, saving space and time.
4. Semantic Split Search: Queries both "Code" and "Docs" separately to ensure documentation doesn't drown out implementation details, then reranks with ColBERT.
5. Global Batching: A producer/consumer pipeline decouples chunking from embedding. Files are chunked concurrently, queued, embedded in fat batches, and written to LanceDB in bulk.
6. Anchor-Only Scans & Batch Deletes: File discovery and stale cleanup hit only anchor rows, and stale/changed paths are removed with a single IN delete to minimize I/O.
7. Structural Boosting: Function/class chunks get a small score boost; test/spec paths are slightly downweighted to bubble up primary definitions first.
8. Role Classification: Detects ORCHESTRATION functions (high complexity, many calls) vs DEFINITION (types/classes) to help agents prioritize where to read.

Configuration

Automatic Repository Isolation

osgrep automatically creates a unique index for each repository based on:

1. Git Remote URL (e.g., github.com/facebook/react → facebook-react)
2. Git Repo without Remote → directory name + hash (e.g., utils-7f8a2b3c)
3. Non-Git Directory → directory name + hash for collision safety

Examples:

cd ~/work/myproject        # Auto-detected: owner-myproject
osgrep "API handlers"

cd ~/personal/utils        # Auto-detected: utils-abc12345
osgrep "helper functions"

Stores are isolated automatically — no manual --store flags needed!

Ignoring Files

osgrep respects both .gitignore and .osgrepignore files when indexing. Create a .osgrepignore file in your repository root to exclude additional files or patterns from indexing.

.osgrepignore syntax:

• Uses the same pattern syntax as .gitignore
• Patterns are relative to the repository root
• Supports glob patterns, negation (!), and directory patterns (/)

Manual Store Management

• View all stores: osgrep list
• Override auto-detection: osgrep --store custom-name "query"
• Clean up old stores: rm -rf ~/.osgrep/data/store-name
• Data location: ~/.osgrep/data

Development

pnpm install
pnpm build        # or pnpm dev
pnpm format       # biome check

Troubleshooting

• Index feels stale? Run osgrep index to refresh.
• Weird results? Run osgrep doctor to verify models.
• Index getting stuck? Run osgrep index --verbose to see which file is being processed.
• Want faster indexing? Keep fallback disabled (default) to skip files without TreeSitter support.
• Need a fresh start? Delete ~/.osgrep/data and ~/.osgrep/meta.json and run osgrep index.

Attribution

osgrep is built upon the foundation of mgrep by MixedBread. We acknowledge and appreciate the original architectural concepts and design decisions that informed this work.

See the NOTICE file for detailed attribution information.

License

Licensed under the Apache License, Version 2.0. See LICENSE and Apache-2.0 for details.