https://github.com/intelligencedev/manifold

Manifold is a platform for enabling workflow automation using AI assistants.
https://github.com/intelligencedev/manifold

artificial-intelligence automation go visual-coding vuejs

Last synced: 11 days ago
JSON representation

Manifold is a platform for enabling workflow automation using AI assistants.

• Host: GitHub
• URL: https://github.com/intelligencedev/manifold
• Owner: intelligencedev
• Created: 2025-02-04T23:28:34.000Z (17 days ago)
• Default Branch: master
• Last Pushed: 2025-02-07T02:56:59.000Z (14 days ago)
• Last Synced: 2025-02-07T03:28:20.019Z (14 days ago)
• Topics: artificial-intelligence, automation, go, visual-coding, vuejs
• Language: Vue
• Homepage: https://intelligence.dev
• Size: 4.24 MB
• Stars: 8
• Watchers: 4
• Forks: 3
• Open Issues: 3
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Manifold

![Manifold](docs/manifold_glsl.png)

Manifold supports OpenAI compatible endpoints such as [llama-server](https://github.com/ggerganov/llama.cpp/tree/master/examples/server) and [Apple's MLX LM server](https://github.com/ml-explore/mlx-examples/blob/main/llms/mlx_lm/SERVER.md)

Combine public and locally hosted models to create powerful workflows that run sequential or concurrent.

## Web QA

![Manifold Web QA](docs/manifold_webqa.png)

## On Demand Tool/Code Generation and Execution

![Manifold Web QA](docs/manifold_code.png)

## DataDog Graph

![Manifold Web QA](docs/manifold_datadog.png)

## Developer Notes

There are various hard coded configurations that need to be broken out into a configuration file the app loads.

For now, developers can change those endpoints and ports via the appropriate component code, or match the hard configuration.

Backend is `localhost:8080` and frontend is `localhost:3000`

We will implement a solution to change the configuration via a `.env` in a future commit.

## Build and Run

```

$ nvm use 20

$ npm run build

$ go build -ldflags="-s -w" -trimpath -o ./dist/manifold main.go

```

The `manifold` binary will contain the compiled frontend and backend.

Run the compiled application:

```

$ cd dist

$ ./manifold

```

### External Dependencies

Jaeger is not required, but we recommend [deploying the container](https://www.jaegertracing.io/docs/1.6/getting-started/#all-in-one-docker-image).

[pgvector](https://github.com/pgvector/pgvector) - SQL and vector store

Requires Node 20. We recommend using [NVM](https://github.com/nvm-sh/nvm) to manage Node environments.

Backend - localhost:8080:

The backend has Open Telemetry support and requires the JAEGER_ENDPOINT (Jaeger) endpoint be set. This does not have to exist so a fake endpoint can be set.

The application will just throw an error when attempting to send telemetry to the endpoint but will still function.

```

$ export JAEGER_ENDPOINT=

$ go mod tidy

$ go run .

```

Frontend - localhost:3000:

```

$ cd frontend

$ nvm use 20

$ npm install

$ npm run dev

```