Files
javadandCursor b93ecf1691 Initial commit of the Kamidso API.
Keep source, config, and deploy files. Leave runtime data, uploads, and logs out of git.

Co-authored-by: Cursor <cursoragent@cursor.com>
2026-10-10 13:43:09 +03:30
..
2026-10-10 13:43:09 +03:30
2026-10-10 13:43:09 +03:30
2026-10-10 13:43:09 +03:30

Kamidso Next — Phase 1

Product architecture with an in-process execution engine. See ARCHITECTURE.md for invariants and boundaries.

Prerequisites

  • Python 3.11+, uv
  • Node.js LTS
  • MongoDB (local or Docker)
  • Model API keys in environment (OPENAI_API_KEY, etc.)

Development

# Terminal 1 — Mongo
docker compose -f docker/kamidso/docker-compose.yml up -d

# Terminal 2 — API (this repository)
export KAMIDSO_ENV=development
export MONGODB_URI=mongodb://127.0.0.1:27017/kamidso
python -m kamidso.cli serve --host 127.0.0.1 --port 5670 --config configs/engine.toml

# Terminal 3 — Frontend repository
npm install && npm run dev

Or from this repository:

make dev

Open http://localhost:5173 — sign in with KAMIDSO_ADMIN_EMAIL / KAMIDSO_ADMIN_PASSWORD.

API boundary

Frontend and product clients use only /api/kamidso/v1/*. DB-GPT routes are internal implementation details.

Configuration

  • Product: configs/kamidso/{development,production,test}.toml
  • Engine: --config / KAMIDSO_ENGINE_CONFIG (default: configs/engine.toml)
  • Secrets: environment variables (MONGODB_URI, KAMIDSO_JWT_SECRET, model keys)
  • Native DB-GPT memory/prompt flags: NATIVE_DBGPT_MEMORY.md

Tests

python -m pytest tests/kamidso -q

Golden benchmark manifest: tests/kamidso/benchmarks/golden.yaml

Repository folders (keep)

These are required for Kamidso development and must not be removed during cleanup:

  • docs/kamidso/ — product docs
  • docker/ — Mongo compose and the Walmart sample CSV
  • configs/ — engine and Kamidso configuration
  • kamidso/ — product API plus execution engine
  • frontend repository — web UI
  • skills/ — analysis skills
  • tests/kamidso/ — product tests

Product entry: kamidso serve, plus the frontend repository. See ARCHITECTURE.md and REPO_LAYOUT.md.

Production layout

Coolify deploys the frontend repository and this repository as two Dockerfile applications, with MongoDB as a separate database resource. See COOLIFY.md.

See REPO_LAYOUT.md for the tree. Summary:

  • Product: kamidso/, configs/kamidso/, tests/kamidso/
  • Engine: kamidso/engine/, configs/engine.toml, skills/
  • Infra compose: docker/kamidso/docker-compose.yml
  • Not in git: data/, logs/, .venv/, .env