README
¶
Self-hosted search across all your team's tools.
Documentation | Discord | Report Bug | Request Feature
Sercha Core connects your team's data sources - GitHub, Google Drive, Notion, Confluence, and more - and provides unified search across all of them. Self-hosted, so your data stays on your infrastructure.
Features
- Connectors - GitHub and LocalFS today, with 12+ more planned
- BM25 search - Full-text search powered by OpenSearch
- Semantic search - Vector search with pgvector and configurable embedding models
- OAuth2 - Connect data sources securely, with JWT session management
- REST API - Full OpenAPI spec with Swagger UI
- Admin UI - Web interface for managing sources, connections, and search
Quick Start
Requires Docker and 4GB RAM.
git clone https://github.com/sercha-oss/sercha-core.git
cd sercha-core/examples/quickstart
docker compose --profile ui up -d
API available at http://localhost:8080, Admin UI at http://localhost:3000.
See the Quickstart Guide for the full walkthrough.
Development
cd examples/dev
docker compose up -d --build # API at localhost:8080
cd ui
npm install && npm run dev # UI at localhost:3000
See examples/dev/ and CONTRIBUTING.md for details.
Documentation
Contributing
See CONTRIBUTING.md for development setup and guidelines, and SECURITY.md for reporting vulnerabilities.
License
Apache 2.0 - see LICENSE.
Directories
¶
| Path | Synopsis |
|---|---|
|
cmd
|
|
|
sercha-core
command
|
|
|
Package docs Code generated by swaggo/swag.
|
Package docs Code generated by swaggo/swag. |
|
internal
|
|
|
adapters/driven/connectors
Package connectors provides process-level shared infrastructure for all external service connectors — token pooling, transport pooling, and the factory that composes them.
|
Package connectors provides process-level shared infrastructure for all external service connectors — token pooling, transport pooling, and the factory that composes them. |
|
adapters/driven/pgvector/migrations
Package migrations holds the pgvector schema migrations as Go code so the embedding dimension (which must appear as a literal in the column type) can be read from PGVECTOR_DIMENSIONS at migrate-time rather than baked into a SQL file at compile-time.
|
Package migrations holds the pgvector schema migrations as Go code so the embedding dimension (which must appear as a literal in the column type) can be read from PGVECTOR_DIMENSIONS at migrate-time rather than baked into a SQL file at compile-time. |
|
adapters/driven/pipeline/stages/chunking
Package chunking provides shared text-splitting helpers for use by both indexing and search-side stages.
|
Package chunking provides shared text-splitting helpers for use by both indexing and search-side stages. |
|
adapters/driven/pipeline/stages/textfilter
Package textfilter provides shared heuristics for detecting non-text content (binary blobs, base64, compressed payloads) in indexable documents.
|
Package textfilter provides shared heuristics for detecting non-text content (binary blobs, base64, compressed payloads) in indexable documents. |
|
adapters/driving/mcp/widgets
Package widgets wires MCP Apps (ext-apps) interactive UI into the Sercha MCP server.
|
Package widgets wires MCP Apps (ext-apps) interactive UI into the Sercha MCP server. |
|
assets
Package assets holds binary assets that multiple adapters embed and serve.
|
Package assets holds binary assets that multiple adapters embed and serve. |
|
Package swagger Code generated by swaggo/swag.
|
Package swagger Code generated by swaggo/swag. |
Click to show internal directories.
Click to hide internal directories.