MCP Terminal Server

# How to contribute We'd love to accept your patches and contributions to this project. ## Before you begin ### Sign our Contributor License Agreement Contributions to this project must be accompanied by a [Contributor License Agreement](https://cla.developers.google.com/about) (CLA). You (or your employer) retain the copyright to your contribution; this simply gives us permission to use and redistribute your contributions as part of the project. If you or your current employer have already signed the Google CLA (even if it was for a different project), you probably don't need to do it again. Visit <https://cla.developers.google.com/> to see your current agreements or to sign a new one. ### Review our community guidelines This project follows [Google's Open Source Community Guidelines](https://opensource.google/conduct/). ## Contribution process ### Code reviews All submissions, including submissions by project members, require review. We use GitHub pull requests for this purpose. Consult [GitHub Help](https://help.github.com/articles/about-pull-requests/) for more information on using pull requests. ## Setup Genkit supports JavaScript, Go, and Python. Before contributing in any of these languages, complete these prerequisites: 1. Install Node.js 20 or later using [nvm](https://nodejs.org/en/download) > **Note:** Node.js v20 or greater is required. Earlier versions may not work properly. 2. Install the Genkit CLI globally: ```bash npm install -g genkit ``` After completing these prerequisites, follow the language-specific setup instructions below. ## Go Guide 1. Install Go 1.24 or later Follow the [official Go installation guide](https://golang.org/doc/install). 2. Configure your AI model Most samples use Google's Gemini model. You'll need to generate an API key at [Google AI Studio](https://aistudio.google.com/app/apikey). Once you have your key, set it in your environment: ```bash export GOOGLE_GENAI_API_KEY=<your-api-key> ``` 3. Run a sample application ```bash cd go/samples # Navigate to samples directory cd <sample-name> # Choose a sample to run go mod tidy # Install Go dependencies genkit start -- go run . # Start the Genkit server and run the application ``` Once running, visit http://localhost:4000 to access the Developer UI. 4. Run tests ```bash cd <test-directory> # Navigate to test directory go test . # Run tests in current directory ``` ## JS Guide ### Install dependencies Run `corepack enable pnpm` to enable pnpm. ``` pnpm i pnpm run setup ``` This will install all dependencies and build all packages. ### Build it ``` pnpm build ``` This will build all packages in this repository. This is recommended the first time you check out a fresh repo or pull new revisions. ### Pack it Pack packages for testing/distribution. Assuming you built everything previously.... ``` pnpm pack:all ``` This will produce tarballs in the `dist` folder. Also `genkit-dist.zip` -- a zip of all the package tarballs. ### Link it You will need the Genkit CLI to run test apps and the Developer UI (this is done for you with `pnpm run setup`): ``` pnpm link-genkit-cli ``` ## Run it ### Run a flow In the `js/testapps` folder you will find some test apps using Genkit. They might contain instructions for how to run them and what setup is necessary. Here's one that requires no setup: ``` cd js/testapps/flow-sample1 genkit start -- tsx --watch src/index.ts ``` Point your browser to http://localhost:4000 ### Run an evaluation We'll be using the `pdfQA` flow for our example. To start, let's make sure we have some context to pull from the vector store. 1. Start the Developer UI ``` cd js/testapps/evals genkit start -- pnpm genkit:dev ``` 2. Click on the `indexPdf` flow in the left nav. 3. Input paths to pdfs you want to index. There's one checked into the directory: ``` "./docs/cat-handbook.pdf" ``` 4. Run an evaluation on a sample dataset checked into the testapp ``` genkit eval:flow pdfQA --input ./data/cat_adoption_questions.json ``` 5. Go to the **Evaluations** tab in the Developer UI to view the evaluation results. ## Code it FYI: `js` and `genkit-tools` are in two separate workspaces. As you make changes you may want to build and test things by running test apps. You can reduce the scope of what you're building by running a specific build command: ``` pnpm build:genkit pnpm build:genkit-tools ``` But you can also go into specific package that you changed and run ``` pnpm build ``` If you are going to be coding for a while then do ``` pnpm build:watch ``` in the package that you're editing. ## Document it If you are contributing to the core Genkit JS SDK (under the `/js` directory), please make sure that all exported members have a valid [JSDoc](https://www.typescriptlang.org/docs/handbook/jsdoc-supported-types.html) associated with them. Use the following command to build and view the new API reference locally. ``` cd js && pnpm build && pnpm typedoc-html && open api-refs-js/index.html ``` ## Send it Once done coding you will want to send a PR. Always do things in a separate branch (by convention name the branch `your-name/feature-something`). Before sending the PR, always run: ``` pnpm format pnpm build ```