Kodit

A code and document intelligence server that indexes Git repositories and provides search through MCP and REST APIs.

AI coding assistants work better when they have access to real examples from your codebase. Kodit indexes your repositories, splits source files into searchable snippets, and serves them to any MCP-compatible assistant. When your assistant needs to write new code, it queries Kodit first and gets back relevant, up-to-date examples drawn from your own projects.

Kodit also handles documents. PDFs, Word files, PowerPoint decks, and spreadsheets are rasterized and indexed so you can search across both code and documentation in one place.

What you get:

Multiple search strategies including BM25 keyword search, semantic vector search, regex grep, and visual document search, each exposed as a separate MCP tool so your assistant picks the right approach for each query
MCP server that works with Claude Code, Cursor, Cline, Kilo Code, and any other MCP-compatible assistant
REST API for programmatic access to search, repositories, enrichments, and indexing status
AI enrichments (optional) including architecture docs, API docs, database schema detection, cookbook examples, and commit summaries, all generated by an LLM
Document intelligence with visual search across PDF pages, Office documents, and images using multimodal embeddings
No external dependencies required for basic operation, with a built-in embedding model and SQLite storage

Quickstart

Docker (recommended)

docker run -p 8080:8080 registry.helix.ml/helix/kodit:latest

This starts Kodit with SQLite storage and a built-in embedding model. No API keys needed.

Pre-built binaries

Download a binary from the releases page, then:

chmod +x kodit
./kodit serve

Verify it works

Open the interactive API docs at http://localhost:8080/docs.

Or index a small repository and run a search:

# Index a repository
curl http://localhost:8080/api/v1/repositories \
  -X POST -H "Content-Type: application/json" \
  -d '{
    "data": {
      "type": "repository",
      "attributes": {
        "remote_uri": "https://gist.github.com/philwinder/7aa38185e20433c04c533f2b28f4e217.git"
      }
    }
  }'

# Check indexing progress
curl http://localhost:8080/api/v1/repositories/1/status

# Search (once indexing is complete)
curl http://localhost:8080/api/v1/search \
  -X POST -H "Content-Type: application/json" \
  -d '{
    "data": {
      "type": "search",
      "attributes": {
        "keywords": ["orders"],
        "text": "code to get all orders"
      }
    }
  }'

Connecting to AI Assistants

Kodit exposes an MCP endpoint at /mcp. Connect your assistant to start using Kodit as a code search tool.

Claude Code

claude mcp add --transport http kodit http://localhost:8080/mcp

Cursor

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "kodit": {
      "url": "http://localhost:8080/mcp"
    }
  }
}

Cline

Add to the MCP Servers configuration (Remote Servers tab):

{
  "mcpServers": {
    "kodit": {
      "autoApprove": [],
      "disabled": false,
      "timeout": 60,
      "type": "streamableHttp",
      "url": "http://localhost:8080/mcp"
    }
  }
}

Kilo Code

Add to the MCP configuration (Edit Project/Global MCP):

{
  "mcpServers": {
    "kodit": {
      "type": "streamable-http",
      "url": "http://localhost:8080/mcp",
      "alwaysAllow": [],
      "disabled": false
    }
  }
}

Replace http://localhost:8080 with your server URL if running remotely.

Encouraging assistants to use Kodit

Some assistants may not call Kodit tools automatically. Add this to your project rules or system prompt to enforce usage:

For every request that involves writing or modifying code, the assistant's first
action must be to call the kodit search MCP tools. Only produce or edit code after
the tool call returns results.

In Cursor, save this as .cursor/rules/kodit.mdc with alwaysApply: true frontmatter.

MCP Tools

Kodit exposes these tools to connected AI assistants:

Tool	Description
`kodit_repositories`	List all indexed repositories
`kodit_semantic_search`	Semantic similarity search across code
`kodit_keyword_search`	BM25 keyword search
`kodit_visual_search`	Search document page images
`kodit_grep`	Regex pattern matching
`kodit_ls`	List files by glob pattern
`kodit_read_resource`	Read file content by URI
`kodit_architecture_docs`	Architecture documentation for a repo
`kodit_api_docs`	Public API documentation
`kodit_database_schema`	Database schema documentation
`kodit_cookbook`	Usage examples and patterns
`kodit_commit_description`	Commit description
`kodit_wiki`	Wiki table of contents
`kodit_wiki_page`	Read a specific wiki page
`kodit_version`	Server version

The enrichment tools (architecture_docs, api_docs, database_schema, cookbook, wiki, commit_description) require an LLM provider to be configured. See Enrichment Providers under Configuration Reference.

Go Library

Kodit can be embedded directly as a Go library. This is how Helix integrates Kodit into its platform.

import "github.com/helixml/kodit"

client, err := kodit.New(
    kodit.WithSQLite(".kodit/data.db"),
)
if err != nil {
    log.Fatal(err)
}
defer client.Close()

// Index a repository
repo, err := client.Repositories.Add(ctx, &service.RepositoryAddParams{
    URL: "https://github.com/kubernetes/kubernetes",
})

// Search
results, err := client.Search.Query(ctx, "create a deployment",
    service.WithLimit(10),
)

for _, snippet := range results.Snippets() {
    fmt.Println(snippet.Path(), snippet.Name())
}

Library options

Option	Description
`WithSQLite(path)`	Use SQLite for storage
`WithPostgresVectorchord(dsn)`	Use PostgreSQL with VectorChord
`WithOpenAI(apiKey)`	OpenAI for embeddings and text
`WithAnthropic(apiKey)`	Anthropic Claude for text (needs separate embedding provider)
`WithTextProvider(p)`	Custom text generation provider
`WithEmbeddingProvider(p)`	Custom embedding provider
`WithRAGPipeline()`	Skip LLM enrichments, index and search only
`WithFullPipeline()`	Require all enrichments (errors without a text provider)
`WithDataDir(dir)`	Data directory (default: `~/.kodit`)
`WithCloneDir(dir)`	Repository clone directory
`WithAPIKeys(keys...)`	API keys for HTTP authentication
`WithWorkerCount(n)`	Number of background workers (default: 1)
`WithPeriodicSyncConfig(cfg)`	Automatic repository sync settings

Search options

Option	Description
`WithSemanticWeight(w)`	Weight for semantic vs keyword search (0.0 to 1.0)
`WithLimit(n)`	Maximum number of results
`WithOffset(n)`	Offset for pagination
`WithLanguages(langs...)`	Filter by programming languages
`WithRepositories(ids...)`	Filter by repository IDs
`WithMinScore(score)`	Minimum score threshold

Go HTTP client

A generated HTTP client is available for calling a remote Kodit server from Go:

go get github.com/helixml/kodit/clients/go

import koditclient "github.com/helixml/kodit/clients/go"

client, err := koditclient.NewClient("https://kodit.example.com")

// List repositories
resp, err := client.GetApiV1Repositories(ctx)

// Search
resp, err := client.PostApiV1SearchMulti(ctx, koditclient.PostApiV1SearchMultiJSONRequestBody{
    TextQuery: "create a deployment",
    TopK:      10,
})

Types are auto-generated from the OpenAPI spec. See the interactive API docs at /docs for the full endpoint list.

Production Deployment

For production use, deploy with PostgreSQL (VectorChord) for scalable vector search and a dedicated LLM provider for enrichments.

Docker Compose

Save this as docker-compose.yaml:

services:
  kodit:
    image: registry.helix.ml/helix/kodit:latest
    ports:
      - "8080:8080"
    command: ["serve"]
    restart: unless-stopped
    depends_on:
      - vectorchord
    environment:
      DATA_DIR: /data
      DB_URL: postgresql://postgres:mysecretpassword@vectorchord:5432/kodit

      # Enrichment LLM (optional, enables AI-generated docs)
      ENRICHMENT_ENDPOINT_BASE_URL: http://ollama:11434
      ENRICHMENT_ENDPOINT_MODEL: ollama/qwen3:1.7b

      # External embedding provider (optional, replaces built-in model)
      # EMBEDDING_ENDPOINT_API_KEY: sk-proj-xxxx
      # EMBEDDING_ENDPOINT_MODEL: openai/text-embedding-3-small

      LOG_LEVEL: INFO
      API_KEYS: ${KODIT_API_KEYS:-}
    volumes:
      - kodit-data:/data

  vectorchord:
    image: tensorchord/vchord-suite:pg17-20250601
    environment:
      POSTGRES_DB: kodit
      POSTGRES_PASSWORD: mysecretpassword
    volumes:
      - vectorchord-data:/var/lib/postgresql/data
    restart: unless-stopped

volumes:
  kodit-data:
  vectorchord-data:

Kubernetes

apiVersion: apps/v1
kind: Deployment
metadata:
  name: vectorchord
spec:
  replicas: 1
  selector:
    matchLabels:
      app: vectorchord
  template:
    metadata:
      labels:
        app: vectorchord
    spec:
      containers:
        - name: vectorchord
          image: tensorchord/vchord-suite:pg17-20250601
          env:
            - name: POSTGRES_DB
              value: kodit
            - name: POSTGRES_PASSWORD
              value: mysecretpassword
          ports:
            - containerPort: 5432
---
apiVersion: v1
kind: Service
metadata:
  name: vectorchord
spec:
  selector:
    app: vectorchord
  ports:
    - port: 5432
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: kodit
spec:
  replicas: 1
  selector:
    matchLabels:
      app: kodit
  template:
    metadata:
      labels:
        app: kodit
    spec:
      containers:
        - name: kodit
          image: registry.helix.ml/helix/kodit:latest # pin to a specific version
          args: ["serve"]
          env: [] # see Configuration Reference for environment variables
          ports:
            - containerPort: 8080
          readinessProbe:
            httpGet:
              path: /
              port: 8080
            initialDelaySeconds: 10
            periodSeconds: 5
---
apiVersion: v1
kind: Service
metadata:
  name: kodit
spec:
  type: LoadBalancer
  selector:
    app: kodit
  ports:
    - port: 8080

Authentication

Set the API_KEYS environment variable to a comma-separated list of keys. Write endpoints (creating repositories, triggering syncs) require a valid key in the Authorization: Bearer <key> header. Search endpoints are open by default.

Configuration Reference

Configuration is done through environment variables. You can also use a .env file:

kodit serve --env-file .env

Server

Variable	Default	Description
`HOST`	`0.0.0.0`	Listen address
`PORT`	`8080`	Listen port
`DATA_DIR`	`~/.kodit`	Data directory for models, clones, and database
`DB_URL`	(empty)	PostgreSQL connection string (uses SQLite if empty)
`LOG_LEVEL`	`INFO`	Logging verbosity: `DEBUG`, `INFO`, `WARN`, `ERROR`
`LOG_FORMAT`	`pretty`	Log format: `pretty` or `json`
`API_KEYS`	(empty)	Comma-separated API keys for write endpoints
`WORKER_COUNT`	`1`	Number of background workers
`SEARCH_LIMIT`	`10`	Default search result limit
`DISABLE_TELEMETRY`	`false`	Disable anonymous usage telemetry

Embedding Provider

These configure an external embedding model. If unset, Kodit uses its built-in model.

Variable	Default	Description
`EMBEDDING_ENDPOINT_BASE_URL`	(empty)	Base URL of embedding service
`EMBEDDING_ENDPOINT_MODEL`	(empty)	Model identifier
`EMBEDDING_ENDPOINT_API_KEY`	(empty)	API key
`EMBEDDING_ENDPOINT_MAX_TOKENS`	`0`	Max tokens per request (0 = provider default)
`EMBEDDING_ENDPOINT_MAX_BATCH_CHARS`	`16000`	Max total characters per embedding batch
`EMBEDDING_ENDPOINT_MAX_BATCH_SIZE`	`1`	Max items per batch
`EMBEDDING_ENDPOINT_TIMEOUT`	`60`	Request timeout in seconds
`EMBEDDING_ENDPOINT_NUM_PARALLEL_TASKS`	`1`	Concurrent embedding requests
`EMBEDDING_ENDPOINT_EXTRA_PARAMS`	(empty)	JSON-encoded extra parameters for the embedding provider
`EMBEDDING_ENDPOINT_QUERY_INSTRUCTION`	(empty)	Instruction prepended to queries for asymmetric retrieval
`EMBEDDING_ENDPOINT_DOCUMENT_INSTRUCTION`	(empty)	Instruction prepended to documents for asymmetric retrieval

Vision Embedding Provider

These configure a remote service for image and text vision embeddings. If unset, Kodit uses its built-in SigLIP2 model.

Variable	Default	Description
`VISION_EMBEDDING_ENDPOINT_BASE_URL`	(empty)	Base URL of vision embedding service
`VISION_EMBEDDING_ENDPOINT_MODEL`	(empty)	Model identifier
`VISION_EMBEDDING_ENDPOINT_API_KEY`	(empty)	API key
`VISION_EMBEDDING_ENDPOINT_MAX_TOKENS`	`0`	Max tokens per request (0 = provider default)
`VISION_EMBEDDING_ENDPOINT_MAX_BATCH_CHARS`	`16000`	Max total characters per embedding batch
`VISION_EMBEDDING_ENDPOINT_MAX_BATCH_SIZE`	`1`	Max items per batch
`VISION_EMBEDDING_ENDPOINT_TIMEOUT`	`60`	Request timeout in seconds
`VISION_EMBEDDING_ENDPOINT_NUM_PARALLEL_TASKS`	`1`	Concurrent vision embedding requests
`VISION_EMBEDDING_ENDPOINT_EXTRA_PARAMS`	(empty)	JSON-encoded extra parameters for the vision embedding provider
`VISION_EMBEDDING_ENDPOINT_QUERY_INSTRUCTION`	(empty)	Instruction prepended to queries for asymmetric retrieval
`VISION_EMBEDDING_ENDPOINT_DOCUMENT_INSTRUCTION`	(empty)	Instruction prepended to documents for asymmetric retrieval

Enrichment Providers

These configure an LLM for generating architecture docs, API docs, database schemas, cookbooks, commit summaries, and wiki pages. Without this, Kodit indexes and searches code but does not generate any AI documentation.

Variable	Default	Description
`ENRICHMENT_ENDPOINT_BASE_URL`	(empty)	Base URL of LLM service
`ENRICHMENT_ENDPOINT_MODEL`	(empty)	Model identifier
`ENRICHMENT_ENDPOINT_API_KEY`	(empty)	API key
`ENRICHMENT_ENDPOINT_NUM_PARALLEL_TASKS`	`1`	Concurrent enrichment requests
`ENRICHMENT_ENDPOINT_TIMEOUT`	`60`	Request timeout in seconds
`ENRICHMENT_ENDPOINT_EXTRA_PARAMS`	(empty)	JSON-encoded extra parameters for the LLM

Enrichment is typically the slowest part of indexing because each enrichment requires a round-trip to the LLM provider. Increase NUM_PARALLEL_TASKS to speed things up, but respect your provider's rate limits. Start low and increase over time.

Provider examples:

# OpenAI
ENRICHMENT_ENDPOINT_BASE_URL=https://api.openai.com/v1
ENRICHMENT_ENDPOINT_MODEL=gpt-4o-mini
ENRICHMENT_ENDPOINT_API_KEY=sk-proj-xxxx

# Ollama (local)
ENRICHMENT_ENDPOINT_BASE_URL=http://localhost:11434
ENRICHMENT_ENDPOINT_MODEL=ollama/qwen3:1.7b

# Helix (private cloud)
ENRICHMENT_ENDPOINT_BASE_URL=https://app.helix.ml/v1
ENRICHMENT_ENDPOINT_MODEL=Qwen/Qwen3-8B
ENRICHMENT_ENDPOINT_API_KEY=your-helix-key

Periodic Sync

Variable	Default	Description
`PERIODIC_SYNC_ENABLED`	`true`	Auto-sync repositories on an interval
`PERIODIC_SYNC_INTERVAL_SECONDS`	`1800`	Sync interval (default: 30 minutes)
`PERIODIC_SYNC_RETRY_ATTEMPTS`	`3`	Retry count on sync failure

Chunking

Variable	Default	Description
`CHUNK_SIZE`	`1500`	Characters per chunk
`CHUNK_OVERLAP`	`200`	Overlap between adjacent chunks
`CHUNK_MIN_SIZE`	`50`	Minimum chunk size

REST API

The full API is documented interactively at /docs on a running Kodit instance. The OpenAPI 3.0 specification is available at /docs/openapi.json.

Key endpoints:

Method	Path	Description
`POST`	`/api/v1/repositories`	Add a repository for indexing
`GET`	`/api/v1/repositories`	List indexed repositories
`GET`	`/api/v1/repositories/{id}/status`	Indexing progress
`POST`	`/api/v1/repositories/{id}/sync`	Trigger a sync
`DELETE`	`/api/v1/repositories/{id}`	Remove a repository
`POST`	`/api/v1/search`	Combined search (keyword + semantic)
`GET`	`/api/v1/search/semantic`	Semantic search only
`GET`	`/api/v1/search/keyword`	Keyword search only
`GET`	`/api/v1/search/visual`	Visual search on document pages
`GET`	`/api/v1/search/grep`	Regex pattern search
`GET`	`/api/v1/search/ls`	List files by glob

All write endpoints require an Authorization: Bearer <key> header when API_KEYS is set.

How Indexing Works

When you add a repository, Kodit runs a pipeline:

Clone the Git repository to local storage
Scan commits, branches, and tags to extract metadata
Extract snippets by splitting source files into overlapping text chunks
Build search indexes with BM25 (keyword) and vector embeddings (semantic)
Generate enrichments (if an LLM provider is configured): architecture docs, API docs, database schemas, cookbook examples, commit summaries, and wiki pages

Kodit tracks which files have changed between syncs and only reprocesses modified content. Repositories sync automatically on a configurable interval (default: every 30 minutes).

Supported sources

Kodit indexes any Git repository accessible via HTTPS, SSH, or the Git protocol. This includes GitHub, GitLab, Bitbucket, Azure DevOps, and self-hosted servers.

Private repositories

Private repositories are supported through personal access tokens or SSH keys:

# HTTPS with token
https://username:token@github.com/username/repo.git

# SSH (ensure your SSH key is configured)
git@github.com:username/repo.git

Privacy

Kodit respects .gitignore and .noindex files. Files matching these patterns are excluded from indexing.

Storage Backends

SQLite (default)

No configuration needed. Kodit creates a SQLite database in the data directory with FTS5 for keyword search and in-process vector storage. Good for single-user and small-team deployments.

PostgreSQL with VectorChord

For larger deployments, use PostgreSQL with the VectorChord extension. This provides scalable vector search and concurrent access. Set the DB_URL environment variable to your connection string.

The recommended Docker image is tensorchord/vchord-suite:pg17-20250601, which bundles PostgreSQL 17 with VectorChord, vchord_bm25, and pg_tokenizer.

Building from Source

git clone https://github.com/helixml/kodit.git
cd kodit
make tools          # Install development tools
make download-model # Download the built-in embedding model
make build          # Build the binary
./bin/kodit version
./bin/kodit serve

Run the tests:

make test                         # All tests
make test PKG=./internal/foo/...  # Specific package
make check                        # Format, vet, lint, and test

Troubleshooting

MCP connection error after restart: If you see No valid session ID provided after restarting the Kodit server, reload the MCP client in your assistant. MCP sessions do not survive server restarts.

No search results: Check that indexing has completed by calling GET /api/v1/repositories/{id}/status. If status shows errors, check the server logs with LOG_LEVEL=DEBUG.

Enrichments not generating: Enrichments require an LLM provider. Check that ENRICHMENT_ENDPOINT_BASE_URL and ENRICHMENT_ENDPOINT_MODEL are set. Without these, Kodit indexes and searches code but does not generate AI documentation.

Telemetry

Kodit collects limited anonymous telemetry (usage metadata only, no user data) to guide development. Disable it with:

DISABLE_TELEMETRY=true

Commercial Support

Helix provides a managed platform built on Kodit with additional features including a management UI, repository browsing, team collaboration, and hosted infrastructure. For commercial support or enterprise integration, contact founders@helix.ml.

Contributing

See CONTRIBUTING.md for guidelines.

License

Apache 2.0

Name		Name	Last commit message	Last commit date
Latest commit History 415 Commits
.claude		.claude
.github		.github
.vscode		.vscode
application		application
charts/kodit		charts/kodit
clients/go		clients/go
cmd		cmd
domain		domain
infrastructure		infrastructure
internal		internal
scripts		scripts
test		test
tools		tools
.air.toml		.air.toml
.dockerignore		.dockerignore
.gitignore		.gitignore
.ort-version		.ort-version
.tokenizers-version		.tokenizers-version
CLAUDE.md		CLAUDE.md
Dockerfile		Dockerfile
LICENSE		LICENSE
Makefile		Makefile
README.md		README.md
docker-compose.dev.yaml		docker-compose.dev.yaml
docker-entrypoint.sh		docker-entrypoint.sh
errors.go		errors.go
go.mod		go.mod
go.sum		go.sum
handlers.go		handlers.go
handlers_test.go		handlers_test.go
kodit.go		kodit.go
kodit_integration_test.go		kodit_integration_test.go
kodit_test.go		kodit_test.go
mcp_server.go		mcp_server.go
mcp_server_test.go		mcp_server_test.go
options.go		options.go
scoped_handler.go		scoped_handler.go
scoped_handler_test.go		scoped_handler_test.go

Folders and files

Latest commit

History

Repository files navigation

Kodit

Quickstart

Docker (recommended)

Pre-built binaries

Verify it works

Connecting to AI Assistants

Claude Code

Cursor

Cline

Kilo Code

Encouraging assistants to use Kodit

MCP Tools

Go Library

Library options

Search options

Go HTTP client

Production Deployment

Docker Compose

Kubernetes

Authentication

Configuration Reference

Server

Embedding Provider

Vision Embedding Provider

Enrichment Providers

Periodic Sync

Chunking

REST API

How Indexing Works

Supported sources

Private repositories

Privacy

Storage Backends

SQLite (default)

PostgreSQL with VectorChord

Building from Source

Troubleshooting

Telemetry

Commercial Support

Contributing

License

About

Topics

Resources

License

Code of conduct

Contributing

Uh oh!

Stars

Watchers

Forks

Releases 111

Uh oh!

Contributors

Uh oh!

Languages