degubase

Degubase

---

---

Why Degubase?

We all have data to wrangle. Spreadsheets often feel too loose, full-blown database engines require setup and hosting, and SaaS platforms demand your email, credit card, and privacy.

Degubase is the anti-platform: a single compiled binary that gives you a complete relational database with a clean, responsive web UI sitting directly on your machine.

• 100% Local & Self-Contained — Backed by SQLite. Everything lives in one file or data directory. Zero Redis, Postgres, or container sprawl.
• Built for Humans & AI Agents — Human-friendly web views with inline editing, plus an instant Model Context Protocol (MCP) server so Claude, Cursor, or your custom agents can read and write data reliably.
• Your Data, Your Rules — Attach Lua scripts that react to row changes, link records across tables, and import/export CSV whenever you want.

---

⚡ Quickstart

Option 1: Docker (Fastest)

Run Degubase with a single command. Data persists automatically in a named volume:

docker run -d \
  --name degubase \
  -p 8080:8080 \
  -v degubase-data:/data \
  -e DEGUBASE_HOST=0.0.0.0 \
  ghcr.io/saintedlama/degubase:latest

Open http://localhost:8080 in your browser.

services:
  degubase:
    image: ghcr.io/saintedlama/degubase:latest
    container_name: degubase
    ports:
      - "8080:8080"
    environment:
      DEGUBASE_HOST: "0.0.0.0"
    volumes:
      - data:/data
      - config:/config
    restart: unless-stopped

volumes:
  data:
  config:

Run with:

docker compose up -d

---

Option 2: Run from Source (Go)

git clone https://github.com/saintedlama/degubase.git
cd degubase
go run ./cmd/server

A SQLite database (degubase.db) is automatically initialized on first run.

💡 Tip: Want to explore Degubase with sample data preloaded? Run make seed-demo (or go run ./cmd/seed_demo) to instantly populate demo workspaces for Task Tracking, CRM, Bug Tracking, Content Planning, and Recipes!

---

✨ Key Features

• Single Binary, Zero Dependencies — Go backend embeds the complete Vue 3 frontend bundle. Run one executable and you're up.
• 20+ Rich Column Types — Designed for real-world information: text, long-text, markdown, email, url, number, currency, percent, rating, date, datetime, checkbox, single-select, multi-select, checklist, file, image (with auto-thumbnails), emoji, symbol, row-link, created-at, and updated-at.
• Linked Records & Hierarchies — Connect rows across tables (assignee ➔ user, invoice ➔ client) or create parent/child trees within the same table (subtask ➔ task, nested pages). Graph traversal API included with circular reference protection.
• Event-Driven Lua Automations — Write lightweight Lua scripts triggered on row create, update, or delete. Compute derived fields, validate inputs, enforce business logic, or make outbound HTTP requests.
• Real-Time UI Sync (SSE) — Server-Sent Events stream database changes live. Boards, grids, and timelines refresh automatically across open browser tabs.
• Row History & Audit Trail — Every edit is tracked with revision IDs and optional user annotations.
• Workspace Security & API Tokens — Optional JWT login, bcrypt password hashing, and scoped SHA-256 bearer tokens for integrations. Or set disable_auth: true for frictionless single-user local use.

---

📊 Five Interactive Views

Switch perspectives on the same table with one click:

View	Best For	Features
Tabular	Data entry & spreadsheet workflows	Inline cell editing, column resizing, bulk operations, row context menus.
Kanban	Workflow & pipeline tracking	Group by any single-select column, drag-and-drop card movement.
Timeline	Editorial calendars & project roadmaps	Swimlanes grouped by channel/owner, date-range scheduling, drag to shift dates.
Matrix	Risk registers & priority grids	2D cross-referencing over X and Y select axes (e.g. Likelihood vs Impact).
Card / Gallery	Portfolios, assets & recipe collections	Visual card layouts, image cover previews, configurable visible fields.

---

🤖 Native MCP Server for AI Agents

Every Degubase workspace exposes a native Model Context Protocol (MCP) server over Server-Sent Events (SSE). This turns Degubase into a persistent, structured memory and tooling layer for AI assistants like Claude Desktop, Cursor, and custom agents.

Quick Setup (Claude Desktop)

1. In Degubase, open your workspace settings, enable the MCP Server toggle, and generate an API Token.
2. Add the server to your claude_desktop_config.json:

{
  "mcpServers": {
    "degubase": {
      "url": "http://localhost:8080/api/workspaces/<WORKSPACE_CODE>/mcp/sse",
      "headers": {
        "Authorization": "Bearer <YOUR_WORKSPACE_TOKEN>"
      }
    }
  }
}

Why Agents Love Degubase:

• Semantic Context Awareness: Workspaces and tables carry descriptive context fields that explain schema semantics to LLMs before they write a query.
• 20+ Dedicated Tools: AI agents can query rows, filter data, inspect linked relations, create tables, update columns, inspect row histories, and run automations programmatically.
• OpenAPI 3.0 Documentation: Auto-generated live specification at /api/openapi.json for custom LLM tooling and API clients.

---

📦 Ready-Made Templates

Degubase includes 16 pre-configured templates that spin up instantly with tailored column schemas and preset default views:

• 🚀 Product & Engineering:
  • Product Roadmap (Now/Next/Later with Kanban & Timeline)
  • Bug Tracker (Markdown reproduction steps, screenshot uploads, triage board & severity matrix)
  • Releases & Changelog (Semantic version tracking, breaking changes, timeline view)
  • Incident Log & Post-Mortem (Outage timestamps, root causes, checklist action items)
  • Goals & Bets (Quarterly bets, key progress metrics, and progress percentages)
  • Feature Requests (User feedback, pain points, demand matrix)
• 💼 Sales & Operations:
  • CRM & Contacts (Contact records, interaction dates, status tracking)
  • Sales Pipeline (Deal stages, currency values, target close dates)
  • Client Projects & Billing (Client deliverables, contract values, invoice status)
  • Launch Checklist (Cross-functional go-live sanity checks)
• 🧠 Strategy & Personal:
  • Risk Register (Pre-built Likelihood × Impact 2D Matrix view)
  • OKRs (Objectives & Key Results) (Measurable targets and initiatives)
  • Decision Journal (Context, alternatives considered, expected vs. actual outcomes)
  • Recipe Book, Reading List, Subscriptions & Renewals
• 🤖 Coding Agents:
  • Agent Orchestration, Evidence Trail, Task Memory Graph

---

🖼️ Showcase

Workflows & Views

Kanban Board View	Timeline Schedule View
Drag-and-drop workflow tracking grouped by column.	Visual swimlane scheduling across date ranges.

Tabular Spreadsheet Grid	Card / Gallery View
Fast spreadsheet editing with rich types and filters.	Custom card layouts with tags, ratings, and status pills.

Pipelines & Details

Sales & Deal Pipeline	Detailed Record View
Deal values, company tags, and stage pipelines.	Markdown rendering, rich attachments, and row actions.

---

⚙️ Configuration & API

Configure via degubase.yaml in your working directory or standard environment variables:

YAML key	Environment variable	Default	Description
port	DEGUBASE_PORT	8080	HTTP listen port
host	DEGUBASE_HOST	127.0.0.1	Network interface to bind (0.0.0.0 for containers/LAN)
data_dir	DEGUBASE_DATA_DIR	data	Directory for SQLite database and uploaded files
jwt_secret	DEGUBASE_JWT_SECRET	(auto)	Secret used to sign authentication JWT tokens
disable_auth	DEGUBASE_DISABLE_AUTH	false	Disables login requirement (auto-authenticates as admin)
automation.http.allowed_hosts	DEGUBASE_AUTOMATION_HTTP_ALLOWED_HOSTS	[]	Hosts or host:port allowed in Lua http.get/post bypassing SSRF checks

Every table and record is accessible via REST with standard JSON payloads:

# 1. Create a workspace
curl -X POST http://localhost:8080/api/workspaces \
  -H "Content-Type: application/json" \
  -d '{"name":"Product Ops","context":"Product development and bug tracking"}'

# 2. Create a table
curl -X POST http://localhost:8080/api/workspaces/PRODUCT_OPS/tables \
  -H "Content-Type: application/json" \
  -d '{"name":"Tasks","context":"Sprint engineering tasks"}'

# 3. Add a column
curl -X POST http://localhost:8080/api/workspaces/PRODUCT_OPS/tables/TASKS/columns \
  -H "Content-Type: application/json" \
  -d '{"name":"Status","type":"single-select","options":{"choices":[["Todo","#808080"],["Done","#22c55e"]]}}'

# 4. Insert a record
curl -X POST http://localhost:8080/api/workspaces/PRODUCT_OPS/tables/TASKS/rows \
  -H "Content-Type: application/json" \
  -d '{"data":{"title":"Implement OAuth","Status":"Todo"}}'

Full interactive documentation is available at http://localhost:8080/api/openapi.json.

Degubase is built with Go and Vue 3 (Vite + Tailwind CSS).

Prerequisites

• Go 1.24+
• pnpm

Common Commands

make dev           # Starts Go backend and Vite UI with hot-module reload
make test          # Runs all Go unit and integration tests
make seed-demo     # Populates local DB with sample workspaces & records
make build         # Builds production binary with embedded UI assets
make lint          # Runs gofmt, go vet, staticcheck, and deadcode checks

---

Tech Stack

Layer	Technology	Rationale
Backend	Go	High-performance, single compiled binary, minimal memory footprint (~25 MB).
Database	SQLite (Pure Go driver)	Zero server configuration, file-based, dependable ACID storage.
Frontend	Vue 3 + Tailwind CSS	Fast reactive UI, smooth drag-and-drop, compact bundle size.
Scripting	GopherLua	Embedded sandbox for custom triggers without external language runtimes.
Agent Interface	Model Context Protocol (MCP)	Direct, standardized AI agent integration over Server-Sent Events.

---

Contributing & Community

Contributions, feature ideas, and bug reports are welcome!

• Check out CONTRIBUTING.md for contribution guidelines and testing conventions.
• Please review our Code of Conduct and Security Policy.

---

License

Degubase is open-source software licensed under the GNU General Public License v3.0.