3.2.4

* 378: fix hook env (#402)

* fix(prepare-commit-msg-hook): update error handling to provide clearer instructions for setting API keys and improve user guidance

* Fix: a bug that causes an error when pushing without setting git remote (#396)

* update deploy commands

* feat(cli): add context flag for providing additional commit message input

* Fix [Bug]: punycode` module is deprecated #426 (#433)

Signed-off-by: Tiger Kaovilai <passawit.kaovilai@gmail.com>

* npm audit fix (#432)

Signed-off-by: Tiger Kaovilai <passawit.kaovilai@gmail.com>

* Feat: Add an option to `Don't push` when there are multiple git remotes (#434)

---------

Co-authored-by: GPT8 <57486732+di-sukharev@users.noreply.github.com>

* feat(engine): add support for MLX AI provider (#437)

---------

Signed-off-by: Emmanuel Ferdman <emmanuelferdman@gmail.com>
Co-authored-by: Emmanuel Ferdman <emmanuelferdman@gmail.com>

* feat(config, engine): add support for Mistral AI provider and engine (#436)

* docs(CONTRIBUTING.md): update `TODO.md` reference (#435)

Signed-off-by: Emmanuel Ferdman <emmanuelferdman@gmail.com>

* feat(config, engine): add support for Mistral AI provider and engine

* ```
feat(package): add mistralai and zod dependencies
```

* fix: recreate package-lock.json with node20

* fix: recreate package-lock.json with node v20.18.1 based on branch dev

---------

Signed-off-by: Emmanuel Ferdman <emmanuelferdman@gmail.com>
Co-authored-by: Emmanuel Ferdman <emmanuelferdman@gmail.com>
Co-authored-by: pedro-valentim <>

---------

Signed-off-by: Tiger Kaovilai <passawit.kaovilai@gmail.com>
Signed-off-by: Emmanuel Ferdman <emmanuelferdman@gmail.com>
Co-authored-by: Takanori Matsumoto <matscube@gmail.com>
Co-authored-by: BILLY Maxime <ozeliurs@gmail.com>
Co-authored-by: Welington Sampaio <welington.sampaio@icloud.com>
Co-authored-by: Tiger Kaovilai <passawit.kaovilai@gmail.com>
Co-authored-by: albi ⚡️ <sigismondi.alberto@gmail.com>
Co-authored-by: Emmanuel Ferdman <emmanuelferdman@gmail.com>
Co-authored-by: Pedro Valentim Silva Leite <18179935+pedro-valentim@users.noreply.github.com>

.github/CONTRIBUTING.md

@@ -18,7 +18,7 @@ To get started, follow these steps:
 1. Clone the project repository locally.
 2. Install dependencies with `npm install`.
 3. Run the project with `npm run dev`.
-4. See [issues](https://github.com/di-sukharev/open-commit/issues) or [TODO.md](../TODO.md) to help the project.
+4. See [issues](https://github.com/di-sukharev/opencommit/issues) or [TODO.md](TODO.md) to help the project.
 
 ## Commit message guidelines
 
@@ -30,7 +30,7 @@ If you encounter any issues while using the project, please report them on the G
 
 ## Contacts
 
-If you have any questions about contributing to the project, please contact by [creating an issue](https://github.com/di-sukharev/open-commit/issues) on the GitHub issue tracker.
+If you have any questions about contributing to the project, please contact by [creating an issue](https://github.com/di-sukharev/opencommit/issues) on the GitHub issue tracker.
 
 ## License
 

.gitignore

@@ -10,4 +10,5 @@ uncaughtExceptions.log
 .vscode
 src/*.json
 .idea
-test.ts
+test.ts
+notes.md

README.md

@@ -2,7 +2,7 @@
   <div>
     <img src=".github/logo-grad.svg" alt="OpenCommit logo"/>
     <h1 align="center">OpenCommit</h1>
-    <h4 align="center">Follow the bird <a href="https://twitter.com/_sukharev_"><img src="https://img.shields.io/twitter/follow/_sukharev_?style=flat&label=_sukharev_&logo=twitter&color=0bf&logoColor=fff" align="center"></a>
+    <h4 align="center">Author <a href="https://twitter.com/_sukharev_"><img src="https://img.shields.io/twitter/follow/_sukharev_?style=flat&label=_sukharev_&logo=twitter&color=0bf&logoColor=fff" align="center"></a>
   </div>
 	<h2>Auto-generate meaningful commits in a second</h2>
 	<p>Killing lame commits with AI 🤯🔫</p>
@@ -16,7 +16,7 @@
     <img src=".github/opencommit-example.png" alt="OpenCommit example"/>
 </div>
 
-All the commits in this repo are authored by OpenCommit — look at [the commits](https://github.com/di-sukharev/opencommit/commit/eae7618d575ee8d2e9fff5de56da79d40c4bc5fc) to see how OpenCommit works. Emojis and long commit descriptions are configurable.
+All the commits in this repo are authored by OpenCommit — look at [the commits](https://github.com/di-sukharev/opencommit/commit/eae7618d575ee8d2e9fff5de56da79d40c4bc5fc) to see how OpenCommit works. Emojis and long commit descriptions are configurable, basically everything is.
 
 ## Setup OpenCommit as a CLI tool
 
@@ -28,36 +28,27 @@ You can use OpenCommit by simply running it via the CLI like this `oco`. 2 secon
    npm install -g opencommit
    ```
 
-   Alternatively run it via `npx opencommit` or `bunx opencommit`
+2. Get your API key from [OpenAI](https://platform.openai.com/account/api-keys) or other supported LLM providers (we support them all). Make sure that you add your OpenAI payment details to your account, so the API works.
-
-   MacOS may ask to run the command with `sudo` when installing a package globally.
-
-2. Get your API key from [OpenAI](https://platform.openai.com/account/api-keys). Make sure that you add your payment details, so the API works.
 
 3. Set the key to OpenCommit config:
 
    ```sh
-   oco config set OCO_OPENAI_API_KEY=<your_api_key>
+   oco config set OCO_API_KEY=<your_api_key>
    ```
 
    Your API key is stored locally in the `~/.opencommit` config file.
 
 ## Usage
 
-You can call OpenCommit directly to generate a commit message for your staged changes:
+You can call OpenCommit with `oco` command to generate a commit message for your staged changes:
-
-```sh
-git add <files...>
-opencommit
-```
-
-You can also use the `oco` shortcut:
 
 ```sh
 git add <files...>
 oco
 ```
 
+Running `git add` is optional, `oco` will do it for you.
+
 ### Running locally with Ollama
 
 You can also run it with local model through ollama:
@@ -68,20 +59,17 @@ You can also run it with local model through ollama:
 
 ```sh
 git add <files...>
-OCO_AI_PROVIDER='ollama' opencommit
+oco config set OCO_AI_PROVIDER='ollama' OCO_MODEL='llama3:8b'
 ```
 
-If you want to use a model other than mistral, you can do so by setting the `OCO_AI_PROVIDER` environment variable as follows:
+Default model is `mistral`.
+
+If you have ollama that is set up in docker/ on another machine with GPUs (not locally), you can change the default endpoint url.
+
+You can do so by setting the `OCO_API_URL` environment variable as follows:
 
 ```sh
-OCO_AI_PROVIDER='ollama/llama3:8b' opencommit
+oco config set OCO_API_URL='http://192.168.1.10:11434/api/chat'
-```
-
-if you have ollama that is set up in docker/ on another machine with GPUs (not locally), you can change the default endpoint url.
-You can do so by setting the `OCO_OLLAMA_API_URL` environment variable as follows:
-
-```sh
-OCO_OLLAMA_API_URL='http://192.168.1.10:11434/api/chat' opencommit
 ```
 
 where 192.168.1.10 is example of endpoint URL, where you have ollama set up.
@@ -95,6 +83,7 @@ There are multiple optional flags that can be used with the `oco` command:
 Link to the GitMoji specification: https://gitmoji.dev/
 
 This flag can only be used if the `OCO_EMOJI` configuration item is set to `true`. This flag allows users to use all emojis in the GitMoji specification, By default, the GitMoji full specification is set to `false`, which only includes 10 emojis (🐛✨📝🚀✅♻️⬆️🔧🌐💡).
+
 This is due to limit the number of tokens sent in each request. However, if you would like to use the full GitMoji specification, you can use the `--fgm` flag.
 
 ```
@@ -116,20 +105,23 @@ oco --yes
 Create a `.env` file and add OpenCommit config variables there like this:
 
 ```env
-OCO_OPENAI_API_KEY=<your OpenAI API token>
+...
+OCO_AI_PROVIDER=<openai (default), anthropic, azure, ollama, gemini, flowise>
+OCO_API_KEY=<your OpenAI API token> // or other LLM provider API token
+OCO_API_URL=<may be used to set proxy path to OpenAI api>
 OCO_TOKENS_MAX_INPUT=<max model token limit (default: 4096)>
 OCO_TOKENS_MAX_OUTPUT=<max response tokens (default: 500)>
-OCO_OPENAI_BASE_PATH=<may be used to set proxy path to OpenAI api>
 OCO_DESCRIPTION=<postface a message with ~3 sentences description of the changes>
 OCO_EMOJI=<boolean, add GitMoji>
-OCO_MODEL=<either 'gpt-4o', 'gpt-4', 'gpt-4-turbo', 'gpt-3.5-turbo' (default), 'gpt-3.5-turbo-0125', 'gpt-4-1106-preview', 'gpt-4-turbo-preview' or 'gpt-4-0125-preview'>
+OCO_MODEL=<either 'gpt-4o', 'gpt-4', 'gpt-4-turbo', 'gpt-3.5-turbo' (default), 'gpt-3.5-turbo-0125', 'gpt-4-1106-preview', 'gpt-4-turbo-preview' or 'gpt-4-0125-preview' or any Anthropic or Ollama model or any string basically, but it should be a valid model name>
 OCO_LANGUAGE=<locale, scroll to the bottom to see options>
 OCO_MESSAGE_TEMPLATE_PLACEHOLDER=<message template placeholder, default: '$msg'>
 OCO_PROMPT_MODULE=<either conventional-commit or @commitlint, default: conventional-commit>
 OCO_ONE_LINE_COMMIT=<one line commit message, default: false>
-OCO_AI_PROVIDER=<anthropic, azure, ollama or ollama/model default ollama model: mistral>
 ```
 
+Global configs are same as local configs, but they are stored in the global `~/.opencommit` config file and set with `oco config set` command, e.g. `oco config set OCO_MODEL=gpt-4o`.
+
 ### Global config for all repos
 
 Local config still has more priority than Global config, but you may set `OCO_MODEL` and `OCO_LOCALE` globally and set local configs for `OCO_EMOJI` and `OCO_DESCRIPTION` per repo which is more convenient.
@@ -137,7 +129,7 @@ Local config still has more priority than Global config, but you may set `OCO_MO
 Simply set any of the variables above like this:
 
 ```sh
-oco config set OCO_MODEL=gpt-4o
+oco config set OCO_MODEL=gpt-4o-mini
 ```
 
 Configure [GitMoji](https://gitmoji.dev/) to preface a message.
@@ -152,42 +144,54 @@ To remove preface emojis:
 oco config set OCO_EMOJI=false
 ```
 
-### Switch to GPT-4 or other models
+Other config options are behaving the same.
 
-By default, OpenCommit uses `gpt-4o` model.
+### Output WHY the changes were done (WIP)
 
-You may switch to GPT-4 which performs better, but costs ~x15 times more 🤠
+You can set the `OCO_WHY` config to `true` to have OpenCommit output a short description of WHY the changes were done after the commit message. Default is `false`.
+
+To make this perform accurate we must store 'what files do' in some kind of an index or embedding and perform a lookup (kinda RAG) for the accurate git commit message. If you feel like building this comment on this ticket https://github.com/di-sukharev/opencommit/issues/398 and let's go from there together.
 
 ```sh
-oco config set OCO_MODEL=gpt-4
+oco config set OCO_WHY=true
+```
+
+### Switch to GPT-4 or other models
+
+By default, OpenCommit uses `gpt-4o-mini` model.
+
+You may switch to gpt-4o which performs better, but costs more 🤠
+
+```sh
+oco config set OCO_MODEL=gpt-4o
 ```
 
 or for as a cheaper option:
 
 ```sh
-oco config set OCO_MODEL=gpt-4o-mini
+oco config set OCO_MODEL=gpt-3.5-turbo
 ```
 
-### Switch to Azure OpenAI
+### Switch to other LLM providers with a custom URL
 
 By default OpenCommit uses [OpenAI](https://openai.com).
 
-You could switch to [Azure OpenAI Service](https://learn.microsoft.com/azure/cognitive-services/openai/)🚀
+You could switch to [Azure OpenAI Service](https://learn.microsoft.com/azure/cognitive-services/openai/) or Flowise or Ollama.
 
 ```sh
-opencommit config set OCO_AI_PROVIDER=azure
+oco config set OCO_AI_PROVIDER=azure OCO_API_KEY=<your_azure_api_key> OCO_API_URL=<your_azure_endpoint>
-```
 
-Of course need to set 'OPENAI_API_KEY'. And also need to set the
+oco config set OCO_AI_PROVIDER=flowise OCO_API_KEY=<your_flowise_api_key> OCO_API_URL=<your_flowise_endpoint>
-'OPENAI_BASE_PATH' for the endpoint and set the deployment name to
+
-'model'.
+oco config set OCO_AI_PROVIDER=ollama OCO_API_KEY=<your_ollama_api_key> OCO_API_URL=<your_ollama_endpoint>
+```
 
 ### Locale configuration
 
 To globally specify the language used to generate commit messages:
 
 ```sh
-# de, German ,Deutsch
+# de, German, Deutsch
 oco config set OCO_LANGUAGE=de
 oco config set OCO_LANGUAGE=German
 oco config set OCO_LANGUAGE=Deutsch
@@ -201,14 +205,16 @@ oco config set OCO_LANGUAGE=française
 The default language setting is **English**
 All available languages are currently listed in the [i18n](https://github.com/di-sukharev/opencommit/tree/master/src/i18n) folder
 
-### Push to git
+### Push to git (gonna be deprecated)
 
-Pushing to git is on by default but if you would like to turn it off just use:
+A prompt for pushing to git is on by default but if you would like to turn it off just use:
 
 ```sh
 oco config set OCO_GITPUSH=false
 ```
 
+and it will exit right after commit is confirmed without asking if you would like to push to remote.
+
 ### Switch to `@commitlint`
 
 OpenCommit allows you to choose the prompt module used to generate commit messages. By default, OpenCommit uses its conventional-commit message generator. However, you can switch to using the `@commitlint` prompt module if you prefer. This option lets you generate commit messages in respect with the local config.
@@ -291,7 +297,7 @@ In our codebase, the implementation of this feature can be found in the followin
 
 ```javascript
 commitMessage = messageTemplate.replace(
-  config?.OCO_MESSAGE_TEMPLATE_PLACEHOLDER,
+  config.OCO_MESSAGE_TEMPLATE_PLACEHOLDER,
   commitMessage
 );
 ```
@@ -348,7 +354,7 @@ Or follow the process of your IDE Source Control feature, when it calls `git com
 
 OpenCommit is now available as a GitHub Action which automatically improves all new commits messages when you push to remote!
 
-This is great if you want to make sure all of the commits in all of your repository branches are meaningful and not lame like `fix1` or `done2`.
+This is great if you want to make sure all commits in all of your repository branches are meaningful and not lame like `fix1` or `done2`.
 
 Create a file `.github/workflows/opencommit.yml` with the contents below:
 
@@ -383,7 +389,7 @@ jobs:
           # set openAI api key in repo actions secrets,
           # for openAI keys go to: https://platform.openai.com/account/api-keys
           # for repo secret go to: <your_repo_url>/settings/secrets/actions
-          OCO_OPENAI_API_KEY: ${{ secrets.OCO_OPENAI_API_KEY }}
+          OCO_API_KEY: ${{ secrets.OCO_API_KEY }}
 
           # customization
           OCO_TOKENS_MAX_INPUT: 4096

package.json

@@ -1,6 +1,6 @@
 {
   "name": "opencommit",
-  "version": "3.0.20",
+  "version": "3.2.4",
   "description": "Auto-generate impressive commits in 1 second. Killing lame commits with AI 🤯🔫",
   "keywords": [
     "git",
@@ -46,8 +46,9 @@
     "dev:gemini": "OCO_AI_PROVIDER='gemini' ts-node ./src/cli.ts",
     "build": "rimraf out && node esbuild.config.js",
     "build:push": "npm run build && git add . && git commit -m 'build' && git push",
-    "deploy": "npm run build:push && git push --tags && npm publish --tag latest",
+    "deploy": "npm publish --tag latest",
-    "deploy:patch": "npm version patch && npm run deploy",
+    "deploy:build": "npm run build:push && git push --tags && npm run deploy",
+    "deploy:patch": "npm version patch && npm run deploy:build",
     "lint": "eslint src --ext ts && tsc --noEmit",
     "format": "prettier --write src",
     "test": "node --no-warnings --experimental-vm-modules $( [ -f ./node_modules/.bin/jest ] && echo ./node_modules/.bin/jest || which jest ) test/unit",
@@ -57,7 +58,8 @@
     "test:unit:docker": "npm run test:docker-build && DOCKER_CONTENT_TRUST=0 docker run --rm oco-test npm run test:unit",
     "test:e2e": "npm run test:e2e:setup && jest test/e2e",
     "test:e2e:setup": "sh test/e2e/setup.sh",
-    "test:e2e:docker": "npm run test:docker-build && DOCKER_CONTENT_TRUST=0 docker run --rm oco-test npm run test:e2e"
+    "test:e2e:docker": "npm run test:docker-build && DOCKER_CONTENT_TRUST=0 docker run --rm oco-test npm run test:e2e",
+    "mlx:start": "OCO_AI_PROVIDER='mlx' node ./out/cli.cjs"
   },
   "devDependencies": {
     "@commitlint/types": "^17.4.4",
@@ -81,14 +83,14 @@
     "@actions/core": "^1.10.0",
     "@actions/exec": "^1.1.1",
     "@actions/github": "^5.1.1",
-    "@azure/openai": "^1.0.0-beta.12",
     "@anthropic-ai/sdk": "^0.19.2",
+    "@azure/openai": "^1.0.0-beta.12",
     "@clack/prompts": "^0.6.1",
     "@dqbd/tiktoken": "^1.0.2",
     "@google/generative-ai": "^0.11.4",
+    "@mistralai/mistralai": "^1.3.5",
     "@octokit/webhooks-schemas": "^6.11.0",
     "@octokit/webhooks-types": "^6.11.0",
-    "ai": "^2.2.14",
     "axios": "^1.3.4",
     "chalk": "^5.2.0",
     "cleye": "^1.3.2",
@@ -97,9 +99,8 @@
     "ignore": "^5.2.4",
     "ini": "^3.0.1",
     "inquirer": "^9.1.4",
-    "openai": "^3.2.1"
+    "openai": "^4.57.0",
-  },
+    "punycode": "^2.3.1",
-  "overrides": {
+    "zod": "^3.23.8"
-    "whatwg-url": "13.0.0"
   }
 }

src/cli.ts

@@ -9,6 +9,7 @@ import { configCommand } from './commands/config';
 import { hookCommand, isHookCalled } from './commands/githook.js';
 import { prepareCommitMessageHook } from './commands/prepare-commit-msg-hook';
 import { checkIsLatestVersion } from './utils/checkIsLatestVersion';
+import { runMigrations } from './migrations/_run.js';
 
 const extraArgs = process.argv.slice(2);
 
@@ -19,6 +20,12 @@ cli(
     commands: [configCommand, hookCommand, commitlintConfigCommand],
     flags: {
       fgm: Boolean,
+      context: {
+        type: String,
+        alias: 'c',
+        description: 'Additional user input context for the commit message',
+        default: ''
+      },
       yes: {
         type: Boolean,
         alias: 'y',
@@ -30,12 +37,13 @@ cli(
     help: { description: packageJSON.description }
   },
   async ({ flags }) => {
+    await runMigrations();
     await checkIsLatestVersion();
 
     if (await isHookCalled()) {
       prepareCommitMessageHook();
     } else {
-      commit(extraArgs, false, flags.fgm, flags.yes);
+      commit(extraArgs, flags.context, false, flags.fgm, flags.yes);
     }
   },
   extraArgs

src/commands/ENUMS.ts

@@ -0,0 +1,5 @@
+export enum COMMANDS {
+  config = 'config',
+  hook = 'hook',
+  commitlint = 'commitlint'
+}

src/commands/commit.ts

@@ -1,6 +1,3 @@
-import chalk from 'chalk';
-import { execa } from 'execa';
-
 import {
   confirm,
   intro,
@@ -10,7 +7,8 @@ import {
   select,
   spinner
 } from '@clack/prompts';
-
+import chalk from 'chalk';
+import { execa } from 'execa';
 import { generateCommitMessageByDiff } from '../generateCommitMessageFromGitDiff';
 import {
   assertGitRepo,
@@ -32,43 +30,53 @@ const getGitRemotes = async () => {
 // Check for the presence of message templates
 const checkMessageTemplate = (extraArgs: string[]): string | false => {
   for (const key in extraArgs) {
-    if (extraArgs[key].includes(config?.OCO_MESSAGE_TEMPLATE_PLACEHOLDER))
+    if (extraArgs[key].includes(config.OCO_MESSAGE_TEMPLATE_PLACEHOLDER))
       return extraArgs[key];
   }
   return false;
 };
 
-const generateCommitMessageFromGitDiff = async (
+interface GenerateCommitMessageFromGitDiffParams {
-  diff: string,
+  diff: string;
-  extraArgs: string[],
+  extraArgs: string[];
-  fullGitMojiSpec: boolean,
+  context?: string;
-  skipCommitConfirmation: boolean
+  fullGitMojiSpec?: boolean;
-): Promise<void> => {
+  skipCommitConfirmation?: boolean;
+}
+
+const generateCommitMessageFromGitDiff = async ({
+  diff,
+  extraArgs,
+  context = '',
+  fullGitMojiSpec = false,
+  skipCommitConfirmation = false
+}: GenerateCommitMessageFromGitDiffParams): Promise<void> => {
   await assertGitRepo();
-  const commitSpinner = spinner();
+  const commitGenerationSpinner = spinner();
-  commitSpinner.start('Generating the commit message');
+  commitGenerationSpinner.start('Generating the commit message');
 
   try {
     let commitMessage = await generateCommitMessageByDiff(
       diff,
-      fullGitMojiSpec
+      fullGitMojiSpec,
+      context
     );
 
     const messageTemplate = checkMessageTemplate(extraArgs);
     if (
-      config?.OCO_MESSAGE_TEMPLATE_PLACEHOLDER &&
+      config.OCO_MESSAGE_TEMPLATE_PLACEHOLDER &&
       typeof messageTemplate === 'string'
     ) {
       const messageTemplateIndex = extraArgs.indexOf(messageTemplate);
       extraArgs.splice(messageTemplateIndex, 1);
 
       commitMessage = messageTemplate.replace(
-        config?.OCO_MESSAGE_TEMPLATE_PLACEHOLDER,
+        config.OCO_MESSAGE_TEMPLATE_PLACEHOLDER,
         commitMessage
       );
     }
 
-    commitSpinner.stop('📝 Commit message generated');
+    commitGenerationSpinner.stop('📝 Commit message generated');
 
     outro(
       `Generated commit message:
@@ -77,27 +85,33 @@ ${commitMessage}
 ${chalk.grey('——————————————————')}`
     );
 
-    const isCommitConfirmedByUser = skipCommitConfirmation || await confirm({
+    const isCommitConfirmedByUser =
-      message: 'Confirm the commit message?'
+      skipCommitConfirmation ||
-    });
+      (await confirm({
+        message: 'Confirm the commit message?'
+      }));
 
-    if (isCommitConfirmedByUser && !isCancel(isCommitConfirmedByUser)) {
+    if (isCancel(isCommitConfirmedByUser)) process.exit(1);
+
+    if (isCommitConfirmedByUser) {
+      const committingChangesSpinner = spinner();
+      committingChangesSpinner.start('Committing the changes');
       const { stdout } = await execa('git', [
         'commit',
         '-m',
         commitMessage,
         ...extraArgs
       ]);
-
+      committingChangesSpinner.stop(
-      outro(`${chalk.green('✔')} Successfully committed`);
+        `${chalk.green('✔')} Successfully committed`
+      );
 
       outro(stdout);
 
       const remotes = await getGitRemotes();
 
       // user isn't pushing, return early
-      if (config?.OCO_GITPUSH === false)
+      if (config.OCO_GITPUSH === false) return;
-          return
 
       if (!remotes.length) {
         const { stdout } = await execa('git', ['push']);
@@ -105,12 +119,14 @@ ${chalk.grey('——————————————————')}`
         process.exit(0);
       }
 
-      if (remotes.length === 1 && config?.OCO_GITPUSH !== true) {
+      if (remotes.length === 1) {
         const isPushConfirmedByUser = await confirm({
           message: 'Do you want to run `git push`?'
         });
 
-        if (isPushConfirmedByUser && !isCancel(isPushConfirmedByUser)) {
+        if (isCancel(isPushConfirmedByUser)) process.exit(1);
+
+        if (isPushConfirmedByUser) {
           const pushSpinner = spinner();
 
           pushSpinner.start(`Running 'git push ${remotes[0]}'`);
@@ -122,8 +138,7 @@ ${chalk.grey('——————————————————')}`
           ]);
 
           pushSpinner.stop(
-            `${chalk.green('✔')} Successfully pushed all commits to ${
+            `${chalk.green('✔')} Successfully pushed all commits to ${remotes[0]
-              remotes[0]
             }`
           );
 
@@ -133,42 +148,51 @@ ${chalk.grey('——————————————————')}`
           process.exit(0);
         }
       } else {
+        const skipOption = `don't push`
         const selectedRemote = (await select({
           message: 'Choose a remote to push to',
-          options: remotes.map((remote) => ({ value: remote, label: remote }))
+          options: [...remotes, skipOption].map((remote) => ({ value: remote, label: remote })),
         })) as string;
 
-        if (!isCancel(selectedRemote)) {
+        if (isCancel(selectedRemote)) process.exit(1);
+
+        if (selectedRemote !== skipOption) {
           const pushSpinner = spinner();
-
+  
           pushSpinner.start(`Running 'git push ${selectedRemote}'`);
-
+  
           const { stdout } = await execa('git', ['push', selectedRemote]);
-
+  
+          if (stdout) outro(stdout);
+  
           pushSpinner.stop(
             `${chalk.green(
               '✔'
-            )} Successfully pushed all commits to ${selectedRemote}`
+            )} successfully pushed all commits to ${selectedRemote}`
           );
-
+        }
-          if (stdout) outro(stdout);
-        } else outro(`${chalk.gray('✖')} process cancelled`);
       }
-    }
+    } else {
-    if (!isCommitConfirmedByUser && !isCancel(isCommitConfirmedByUser)) {
       const regenerateMessage = await confirm({
-        message: 'Do you want to regenerate the message ?'
+        message: 'Do you want to regenerate the message?'
       });
-      if (regenerateMessage && !isCancel(isCommitConfirmedByUser)) {
+
-        await generateCommitMessageFromGitDiff(
+      if (isCancel(regenerateMessage)) process.exit(1);
+
+      if (regenerateMessage) {
+        await generateCommitMessageFromGitDiff({
           diff,
           extraArgs,
           fullGitMojiSpec
-        )
+        });
       }
     }
   } catch (error) {
-    commitSpinner.stop('📝 Commit message generated');
+    commitGenerationSpinner.stop(
+      `${chalk.red('✖')} Failed to generate the commit message`
+    );
+
+    console.log(error);
 
     const err = error as Error;
     outro(`${chalk.red('✖')} ${err?.message || err}`);
@@ -178,6 +202,7 @@ ${chalk.grey('——————————————————')}`
 
 export async function commit(
   extraArgs: string[] = [],
+  context: string = '',
   isStageAllFlag: Boolean = false,
   fullGitMojiSpec: boolean = false,
   skipCommitConfirmation: boolean = false
@@ -216,11 +241,10 @@ export async function commit(
       message: 'Do you want to stage all files and generate commit message?'
     });
 
-    if (
+    if (isCancel(isStageAllAndCommitConfirmedByUser)) process.exit(1);
-      isStageAllAndCommitConfirmedByUser &&
+
-      !isCancel(isStageAllAndCommitConfirmedByUser)
+    if (isStageAllAndCommitConfirmedByUser) {
-    ) {
+      await commit(extraArgs, context, true, fullGitMojiSpec);
-      await commit(extraArgs, true, fullGitMojiSpec);
       process.exit(1);
     }
 
@@ -238,7 +262,7 @@ export async function commit(
       await gitAdd({ files });
     }
 
-    await commit(extraArgs, false, fullGitMojiSpec);
+    await commit(extraArgs, context, false, fullGitMojiSpec);
     process.exit(1);
   }
 
@@ -249,12 +273,13 @@ export async function commit(
   );
 
   const [, generateCommitError] = await trytm(
-    generateCommitMessageFromGitDiff(
+    generateCommitMessageFromGitDiff({
-      await getDiff({ files: stagedFiles }),
+      diff: await getDiff({ files: stagedFiles }),
       extraArgs,
+      context,
       fullGitMojiSpec,
       skipCommitConfirmation
-    )
+    })
   );
 
   if (generateCommitError) {

src/commands/commitlint.ts

@@ -1,11 +1,9 @@
+import { intro, outro } from '@clack/prompts';
 import chalk from 'chalk';
 import { command } from 'cleye';
-
-import { intro, outro } from '@clack/prompts';
-
-import { COMMANDS } from '../CommandsEnum';
 import { configureCommitlintIntegration } from '../modules/commitlint/config';
 import { getCommitlintLLMConfig } from '../modules/commitlint/utils';
+import { COMMANDS } from './ENUMS';
 
 export enum CONFIG_MODES {
   get = 'get',
@@ -25,7 +23,7 @@ export const commitlintConfigCommand = command(
       if (mode === CONFIG_MODES.get) {
         const commitLintConfig = await getCommitlintLLMConfig();
 
-        outro(commitLintConfig.toString());
+        outro(JSON.stringify(commitLintConfig, null, 2));
 
         return;
       }

src/commands/config.ts

@@ -1,3 +1,4 @@
+import { intro, outro } from '@clack/prompts';
 import chalk from 'chalk';
 import { command } from 'cleye';
 import * as dotenv from 'dotenv';
@@ -5,37 +6,26 @@ import { existsSync, readFileSync, writeFileSync } from 'fs';
 import { parse as iniParse, stringify as iniStringify } from 'ini';
 import { homedir } from 'os';
 import { join as pathJoin, resolve as pathResolve } from 'path';
-
+import { COMMANDS } from './ENUMS';
-import { intro, outro } from '@clack/prompts';
-
-import { COMMANDS } from '../CommandsEnum';
-import { getI18nLocal } from '../i18n';
 import { TEST_MOCK_TYPES } from '../engine/testAi';
+import { getI18nLocal, i18n } from '../i18n';
 
 export enum CONFIG_KEYS {
-  OCO_OPENAI_API_KEY = 'OCO_OPENAI_API_KEY',
+  OCO_API_KEY = 'OCO_API_KEY',
-  OCO_ANTHROPIC_API_KEY = 'OCO_ANTHROPIC_API_KEY',
-  OCO_AZURE_API_KEY = 'OCO_AZURE_API_KEY',
-  OCO_GEMINI_API_KEY = 'OCO_GEMINI_API_KEY',
-  OCO_GEMINI_BASE_PATH = 'OCO_GEMINI_BASE_PATH',
   OCO_TOKENS_MAX_INPUT = 'OCO_TOKENS_MAX_INPUT',
   OCO_TOKENS_MAX_OUTPUT = 'OCO_TOKENS_MAX_OUTPUT',
-  OCO_OPENAI_BASE_PATH = 'OCO_OPENAI_BASE_PATH',
   OCO_DESCRIPTION = 'OCO_DESCRIPTION',
   OCO_EMOJI = 'OCO_EMOJI',
   OCO_MODEL = 'OCO_MODEL',
   OCO_LANGUAGE = 'OCO_LANGUAGE',
+  OCO_WHY = 'OCO_WHY',
   OCO_MESSAGE_TEMPLATE_PLACEHOLDER = 'OCO_MESSAGE_TEMPLATE_PLACEHOLDER',
   OCO_PROMPT_MODULE = 'OCO_PROMPT_MODULE',
   OCO_AI_PROVIDER = 'OCO_AI_PROVIDER',
-  OCO_GITPUSH = 'OCO_GITPUSH',
   OCO_ONE_LINE_COMMIT = 'OCO_ONE_LINE_COMMIT',
-  OCO_AZURE_ENDPOINT = 'OCO_AZURE_ENDPOINT',
   OCO_TEST_MOCK_TYPE = 'OCO_TEST_MOCK_TYPE',
   OCO_API_URL = 'OCO_API_URL',
-  OCO_OLLAMA_API_URL = 'OCO_OLLAMA_API_URL',
+  OCO_GITPUSH = 'OCO_GITPUSH' // todo: deprecate
-  OCO_FLOWISE_ENDPOINT = 'OCO_FLOWISE_ENDPOINT',
-  OCO_FLOWISE_API_KEY = 'OCO_FLOWISE_API_KEY'
 }
 
 export enum CONFIG_MODES {
@@ -86,6 +76,58 @@ export const MODEL_LIST = {
     'gemini-1.0-pro',
     'gemini-pro-vision',
     'text-embedding-004'
+  ],
+
+  groq: [
+    'llama3-70b-8192', // Meta Llama 3 70B (default one, no daily token limit and 14 400 reqs/day)
+    'llama3-8b-8192', // Meta Llama 3 8B
+    'llama-guard-3-8b', // Llama Guard 3 8B
+    'llama-3.1-8b-instant', // Llama 3.1 8B (Preview)
+    'llama-3.1-70b-versatile', // Llama 3.1 70B (Preview)
+    'gemma-7b-it', // Gemma 7B
+    'gemma2-9b-it' // Gemma 2 9B
+  ],
+
+  mistral: [
+    'ministral-3b-2410',
+    'ministral-3b-latest',
+    'ministral-8b-2410',
+    'ministral-8b-latest',
+    'open-mistral-7b',
+    'mistral-tiny',
+    'mistral-tiny-2312',
+    'open-mistral-nemo',
+    'open-mistral-nemo-2407',
+    'mistral-tiny-2407',
+    'mistral-tiny-latest',
+    'open-mixtral-8x7b',
+    'mistral-small',
+    'mistral-small-2312',
+    'open-mixtral-8x22b',
+    'open-mixtral-8x22b-2404',
+    'mistral-small-2402',
+    'mistral-small-2409',
+    'mistral-small-latest',
+    'mistral-medium-2312',
+    'mistral-medium',
+    'mistral-medium-latest',
+    'mistral-large-2402',
+    'mistral-large-2407',
+    'mistral-large-2411',
+    'mistral-large-latest',
+    'pixtral-large-2411',
+    'pixtral-large-latest',
+    'codestral-2405',
+    'codestral-latest',
+    'codestral-mamba-2407',
+    'open-codestral-mamba',
+    'codestral-mamba-latest',
+    'pixtral-12b-2409',
+    'pixtral-12b',
+    'pixtral-12b-latest',
+    'mistral-embed',
+    'mistral-moderation-2411',
+    'mistral-moderation-latest',
   ]
 };
 
@@ -93,18 +135,24 @@ const getDefaultModel = (provider: string | undefined): string => {
   switch (provider) {
     case 'ollama':
       return '';
+    case 'mlx':
+      return '';
     case 'anthropic':
       return MODEL_LIST.anthropic[0];
     case 'gemini':
       return MODEL_LIST.gemini[0];
+    case 'groq':
+      return MODEL_LIST.groq[0];
+    case 'mistral':
+      return MODEL_LIST.mistral[0];
     default:
       return MODEL_LIST.openai[0];
   }
 };
 
 export enum DEFAULT_TOKEN_LIMITS {
-  DEFAULT_MAX_TOKENS_INPUT = 4096,
+  DEFAULT_MAX_TOKENS_INPUT = 40960,
-  DEFAULT_MAX_TOKENS_OUTPUT = 500
+  DEFAULT_MAX_TOKENS_OUTPUT = 4096
 }
 
 const validateConfig = (
@@ -113,8 +161,10 @@ const validateConfig = (
   validationMessage: string
 ) => {
   if (!condition) {
+    outro(`${chalk.red('✖')} wrong value for ${key}: ${validationMessage}.`);
+
     outro(
-      `${chalk.red('✖')} Unsupported config key ${key}: ${validationMessage}`
+      'For more help refer to docs https://github.com/di-sukharev/opencommit'
     );
 
     process.exit(1);
@@ -122,76 +172,19 @@ const validateConfig = (
 };
 
 export const configValidators = {
-  [CONFIG_KEYS.OCO_OPENAI_API_KEY](value: any, config: any = {}) {
+  [CONFIG_KEYS.OCO_API_KEY](value: any, config: any = {}) {
-    if (config.OCO_AI_PROVIDER == 'gemini') return value;
+    if (config.OCO_AI_PROVIDER !== 'openai') return value;
 
-    //need api key unless running locally with ollama
     validateConfig(
-      'OpenAI API_KEY',
+      'OCO_API_KEY',
-      value ||
+      typeof value === 'string' && value.length > 0,
-        config.OCO_ANTHROPIC_API_KEY ||
+      'Empty value is not allowed'
-        config.OCO_AI_PROVIDER.startsWith('ollama') ||
-        config.OCO_AZURE_API_KEY ||
-        config.OCO_AI_PROVIDER == 'test' ||
-        config.OCO_AI_PROVIDER == 'flowise',
-      'You need to provide an OpenAI/Anthropic/Azure or other provider API key via `oco config set OCO_OPENAI_API_KEY=your_key`, for help refer to docs https://github.com/di-sukharev/opencommit'
-    );
-    validateConfig(
-      CONFIG_KEYS.OCO_OPENAI_API_KEY,
-      value.startsWith('sk-') || config.OCO_AI_PROVIDER != 'openai',
-      'Must start with "sk-" for openai provider'
     );
 
-    return value;
-  },
-
-  [CONFIG_KEYS.OCO_AZURE_API_KEY](value: any, config: any = {}) {
     validateConfig(
-      'ANTHROPIC_API_KEY',
+      'OCO_API_KEY',
-      value ||
+      value,
-        config.OCO_OPENAI_API_KEY ||
+      'You need to provide the OCO_API_KEY when OCO_AI_PROVIDER set to "openai" (default) or "ollama" or "mlx" or "azure" or "gemini" or "flowise" or "anthropic". Run `oco config set OCO_API_KEY=your_key OCO_AI_PROVIDER=openai`'
-        config.OCO_AZURE_API_KEY ||
-        config.OCO_AI_PROVIDER == 'ollama' ||
-        config.OCO_AI_PROVIDER == 'test' ||
-        config.OCO_AI_PROVIDER == 'flowise',
-      'You need to provide an OpenAI/Anthropic/Azure API key'
-    );
-
-    return value;
-  },
-
-  [CONFIG_KEYS.OCO_GEMINI_API_KEY](value: any, config: any = {}) {
-    // only need to check for gemini api key if using gemini
-    if (config.OCO_AI_PROVIDER != 'gemini') return value;
-
-    validateConfig(
-      'Gemini API Key',
-      value || config.OCO_GEMINI_API_KEY || config.OCO_AI_PROVIDER == 'test',
-      'You need to provide an Gemini API key'
-    );
-
-    return value;
-  },
-
-  [CONFIG_KEYS.OCO_ANTHROPIC_API_KEY](value: any, config: any = {}) {
-    validateConfig(
-      'ANTHROPIC_API_KEY',
-      value ||
-        config.OCO_OPENAI_API_KEY ||
-        config.OCO_AI_PROVIDER == 'ollama' ||
-        config.OCO_AI_PROVIDER == 'test' ||
-        config.OCO_AI_PROVIDER == 'flowise',
-      'You need to provide an OpenAI/Anthropic API key'
-    );
-
-    return value;
-  },
-
-  [CONFIG_KEYS.OCO_FLOWISE_API_KEY](value: any, config: any = {}) {
-    validateConfig(
-      CONFIG_KEYS.OCO_FLOWISE_API_KEY,
-      value || config.OCO_AI_PROVIDER != 'flowise',
-      'You need to provide a flowise API key'
     );
 
     return value;
@@ -201,25 +194,17 @@ export const configValidators = {
     validateConfig(
       CONFIG_KEYS.OCO_DESCRIPTION,
       typeof value === 'boolean',
-      'Must be true or false'
+      'Must be boolean: true or false'
     );
 
     return value;
   },
 
   [CONFIG_KEYS.OCO_TOKENS_MAX_INPUT](value: any) {
-    // If the value is a string, convert it to a number.
+    value = parseInt(value);
-    if (typeof value === 'string') {
-      value = parseInt(value);
-      validateConfig(
-        CONFIG_KEYS.OCO_TOKENS_MAX_INPUT,
-        !isNaN(value),
-        'Must be a number'
-      );
-    }
     validateConfig(
       CONFIG_KEYS.OCO_TOKENS_MAX_INPUT,
-      value ? typeof value === 'number' : undefined,
+      !isNaN(value),
       'Must be a number'
     );
 
@@ -227,18 +212,10 @@ export const configValidators = {
   },
 
   [CONFIG_KEYS.OCO_TOKENS_MAX_OUTPUT](value: any) {
-    // If the value is a string, convert it to a number.
+    value = parseInt(value);
-    if (typeof value === 'string') {
-      value = parseInt(value);
-      validateConfig(
-        CONFIG_KEYS.OCO_TOKENS_MAX_OUTPUT,
-        !isNaN(value),
-        'Must be a number'
-      );
-    }
     validateConfig(
       CONFIG_KEYS.OCO_TOKENS_MAX_OUTPUT,
-      value ? typeof value === 'number' : undefined,
+      !isNaN(value),
       'Must be a number'
     );
 
@@ -249,26 +226,29 @@ export const configValidators = {
     validateConfig(
       CONFIG_KEYS.OCO_EMOJI,
       typeof value === 'boolean',
-      'Must be true or false'
+      'Must be boolean: true or false'
     );
 
     return value;
   },
 
   [CONFIG_KEYS.OCO_LANGUAGE](value: any) {
+    const supportedLanguages = Object.keys(i18n);
+
     validateConfig(
       CONFIG_KEYS.OCO_LANGUAGE,
       getI18nLocal(value),
-      `${value} is not supported yet`
+      `${value} is not supported yet. Supported languages: ${supportedLanguages}`
     );
+
     return getI18nLocal(value);
   },
 
-  [CONFIG_KEYS.OCO_OPENAI_BASE_PATH](value: any) {
+  [CONFIG_KEYS.OCO_API_URL](value: any) {
     validateConfig(
-      CONFIG_KEYS.OCO_OPENAI_BASE_PATH,
+      CONFIG_KEYS.OCO_API_URL,
       typeof value === 'string',
-      'Must be string'
+      `${value} is not a valid URL. It should start with 'http://' or 'https://'.`
     );
     return value;
   },
@@ -304,6 +284,7 @@ export const configValidators = {
     return value;
   },
 
+  // todo: deprecate
   [CONFIG_KEYS.OCO_GITPUSH](value: any) {
     validateConfig(
       CONFIG_KEYS.OCO_GITPUSH,
@@ -314,19 +295,23 @@ export const configValidators = {
   },
 
   [CONFIG_KEYS.OCO_AI_PROVIDER](value: any) {
+    if (!value) value = 'openai';
+
     validateConfig(
       CONFIG_KEYS.OCO_AI_PROVIDER,
       [
-        '',
         'openai',
+        'mistral',
         'anthropic',
         'gemini',
         'azure',
         'test',
-        'flowise'
+        'flowise',
+        'groq'
       ].includes(value) || value.startsWith('ollama'),
-      `${value} is not supported yet, use 'ollama', 'anthropic', 'azure', 'gemini', 'flowise' or 'openai' (default)`
+      `${value} is not supported yet, use 'ollama', 'mlx', 'anthropic', 'azure', 'gemini', 'flowise', 'mistral' or 'openai' (default)`
     );
+
     return value;
   },
 
@@ -340,26 +325,6 @@ export const configValidators = {
     return value;
   },
 
-  [CONFIG_KEYS.OCO_AZURE_ENDPOINT](value: any) {
-    validateConfig(
-      CONFIG_KEYS.OCO_AZURE_ENDPOINT,
-      value.includes('openai.azure.com'),
-      'Must be in format "https://<resource name>.openai.azure.com/"'
-    );
-
-    return value;
-  },
-
-  [CONFIG_KEYS.OCO_FLOWISE_ENDPOINT](value: any) {
-    validateConfig(
-      CONFIG_KEYS.OCO_FLOWISE_ENDPOINT,
-      typeof value === 'string' && value.includes(':'),
-      'Value must be string and should include both I.P. and port number' // Considering the possibility of DNS lookup or feeding the I.P. explicitely, there is no pattern to verify, except a column for the port number
-    );
-
-    return value;
-  },
-
   [CONFIG_KEYS.OCO_TEST_MOCK_TYPE](value: any) {
     validateConfig(
       CONFIG_KEYS.OCO_TEST_MOCK_TYPE,
@@ -371,83 +336,64 @@ export const configValidators = {
     return value;
   },
 
-  [CONFIG_KEYS.OCO_OLLAMA_API_URL](value: any) {
+  [CONFIG_KEYS.OCO_WHY](value: any) {
-    // add simple api validator
     validateConfig(
-      CONFIG_KEYS.OCO_API_URL,
+      CONFIG_KEYS.OCO_WHY,
-      typeof value === 'string' && value.startsWith('http'),
+      typeof value === 'boolean',
-      `${value} is not a valid URL`
+      'Must be true or false'
     );
     return value;
   }
 };
 
+export enum OCO_AI_PROVIDER_ENUM {
+  OLLAMA = 'ollama',
+  OPENAI = 'openai',
+  ANTHROPIC = 'anthropic',
+  GEMINI = 'gemini',
+  AZURE = 'azure',
+  TEST = 'test',
+  FLOWISE = 'flowise',
+  GROQ = 'groq',
+  MISTRAL = 'mistral',
+  MLX = 'mlx'
+}
+
 export type ConfigType = {
-  [key in CONFIG_KEYS]?: any;
+  [CONFIG_KEYS.OCO_API_KEY]?: string;
+  [CONFIG_KEYS.OCO_TOKENS_MAX_INPUT]: number;
+  [CONFIG_KEYS.OCO_TOKENS_MAX_OUTPUT]: number;
+  [CONFIG_KEYS.OCO_API_URL]?: string;
+  [CONFIG_KEYS.OCO_DESCRIPTION]: boolean;
+  [CONFIG_KEYS.OCO_EMOJI]: boolean;
+  [CONFIG_KEYS.OCO_WHY]: boolean;
+  [CONFIG_KEYS.OCO_MODEL]: string;
+  [CONFIG_KEYS.OCO_LANGUAGE]: string;
+  [CONFIG_KEYS.OCO_MESSAGE_TEMPLATE_PLACEHOLDER]: string;
+  [CONFIG_KEYS.OCO_PROMPT_MODULE]: OCO_PROMPT_MODULE_ENUM;
+  [CONFIG_KEYS.OCO_AI_PROVIDER]: OCO_AI_PROVIDER_ENUM;
+  [CONFIG_KEYS.OCO_GITPUSH]: boolean;
+  [CONFIG_KEYS.OCO_ONE_LINE_COMMIT]: boolean;
+  [CONFIG_KEYS.OCO_TEST_MOCK_TYPE]: string;
 };
 
-const defaultConfigPath = pathJoin(homedir(), '.opencommit');
+export const defaultConfigPath = pathJoin(homedir(), '.opencommit');
-const defaultEnvPath = pathResolve(process.cwd(), '.env');
+export const defaultEnvPath = pathResolve(process.cwd(), '.env');
-export const getConfig = ({
-  configPath = defaultConfigPath,
-  envPath = defaultEnvPath
-}: {
-  configPath?: string;
-  envPath?: string;
-} = {}): ConfigType | null => {
-  dotenv.config({ path: envPath });
-  const configFromEnv = {
-    OCO_OPENAI_API_KEY: process.env.OCO_OPENAI_API_KEY,
-    OCO_ANTHROPIC_API_KEY: process.env.OCO_ANTHROPIC_API_KEY,
-    OCO_AZURE_API_KEY: process.env.OCO_AZURE_API_KEY,
-    OCO_GEMINI_API_KEY: process.env.OCO_GEMINI_API_KEY,
-    OCO_TOKENS_MAX_INPUT: process.env.OCO_TOKENS_MAX_INPUT
-      ? Number(process.env.OCO_TOKENS_MAX_INPUT)
-      : undefined,
-    OCO_TOKENS_MAX_OUTPUT: process.env.OCO_TOKENS_MAX_OUTPUT
-      ? Number(process.env.OCO_TOKENS_MAX_OUTPUT)
-      : undefined,
-    OCO_OPENAI_BASE_PATH: process.env.OCO_OPENAI_BASE_PATH,
-    OCO_GEMINI_BASE_PATH: process.env.OCO_GEMINI_BASE_PATH,
-    OCO_DESCRIPTION: process.env.OCO_DESCRIPTION === 'true' ? true : false,
-    OCO_EMOJI: process.env.OCO_EMOJI === 'true' ? true : false,
-    OCO_MODEL:
-      process.env.OCO_MODEL || getDefaultModel(process.env.OCO_AI_PROVIDER),
-    OCO_LANGUAGE: process.env.OCO_LANGUAGE || 'en',
-    OCO_MESSAGE_TEMPLATE_PLACEHOLDER:
-      process.env.OCO_MESSAGE_TEMPLATE_PLACEHOLDER || '$msg',
-    OCO_PROMPT_MODULE: process.env.OCO_PROMPT_MODULE || 'conventional-commit',
-    OCO_AI_PROVIDER: process.env.OCO_AI_PROVIDER || 'openai',
-    OCO_GITPUSH: process.env.OCO_GITPUSH === 'false' ? false : true,
-    OCO_ONE_LINE_COMMIT:
-      process.env.OCO_ONE_LINE_COMMIT === 'true' ? true : false,
-    OCO_AZURE_ENDPOINT: process.env.OCO_AZURE_ENDPOINT || undefined,
-    OCO_TEST_MOCK_TYPE: process.env.OCO_TEST_MOCK_TYPE || 'commit-message',
-    OCO_FLOWISE_ENDPOINT: process.env.OCO_FLOWISE_ENDPOINT || ':',
-    OCO_FLOWISE_API_KEY: process.env.OCO_FLOWISE_API_KEY || undefined,
-    OCO_OLLAMA_API_URL: process.env.OCO_OLLAMA_API_URL || undefined
-  };
-  const configExists = existsSync(configPath);
-  if (!configExists) return configFromEnv;
 
-  const configFile = readFileSync(configPath, 'utf8');
+const assertConfigsAreValid = (config: Record<string, any>) => {
-  const config = iniParse(configFile);
+  for (const [key, value] of Object.entries(config)) {
+    if (!value) continue;
 
-  for (const configKey of Object.keys(config)) {
+    if (typeof value === 'string' && ['null', 'undefined'].includes(value)) {
-    if (['null', 'undefined'].includes(config[configKey])) {
+      config[key] = undefined;
-      config[configKey] = undefined;
       continue;
     }
-    try {
-      const validator = configValidators[configKey as CONFIG_KEYS];
-      const validValue = validator(
-        config[configKey] ?? configFromEnv[configKey as CONFIG_KEYS],
-        config
-      );
 
-      config[configKey] = validValue;
+    try {
+      const validate = configValidators[key as CONFIG_KEYS];
+      validate(value, config);
     } catch (error) {
-      outro(`Unknown '${configKey}' config option or missing validator.`);
+      outro(`Unknown '${key}' config option or missing validator.`);
       outro(
         `Manually fix the '.env' file or global '~/.opencommit' config file.`
       );
@@ -455,37 +401,187 @@ export const getConfig = ({
       process.exit(1);
     }
   }
+};
 
-  return config;
+enum OCO_PROMPT_MODULE_ENUM {
+  CONVENTIONAL_COMMIT = 'conventional-commit',
+  COMMITLINT = '@commitlint'
+}
+
+export const DEFAULT_CONFIG = {
+  OCO_TOKENS_MAX_INPUT: DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_INPUT,
+  OCO_TOKENS_MAX_OUTPUT: DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_OUTPUT,
+  OCO_DESCRIPTION: false,
+  OCO_EMOJI: false,
+  OCO_MODEL: getDefaultModel('openai'),
+  OCO_LANGUAGE: 'en',
+  OCO_MESSAGE_TEMPLATE_PLACEHOLDER: '$msg',
+  OCO_PROMPT_MODULE: OCO_PROMPT_MODULE_ENUM.CONVENTIONAL_COMMIT,
+  OCO_AI_PROVIDER: OCO_AI_PROVIDER_ENUM.OPENAI,
+  OCO_ONE_LINE_COMMIT: false,
+  OCO_TEST_MOCK_TYPE: 'commit-message',
+  OCO_WHY: false,
+  OCO_GITPUSH: true // todo: deprecate
+};
+
+const initGlobalConfig = (configPath: string = defaultConfigPath) => {
+  writeFileSync(configPath, iniStringify(DEFAULT_CONFIG), 'utf8');
+  return DEFAULT_CONFIG;
+};
+
+const parseConfigVarValue = (value?: any) => {
+  try {
+    return JSON.parse(value);
+  } catch (error) {
+    return value;
+  }
+};
+
+const getEnvConfig = (envPath: string) => {
+  dotenv.config({ path: envPath });
+
+  return {
+    OCO_MODEL: process.env.OCO_MODEL,
+    OCO_API_URL: process.env.OCO_API_URL,
+    OCO_API_KEY: process.env.OCO_API_KEY,
+    OCO_AI_PROVIDER: process.env.OCO_AI_PROVIDER as OCO_AI_PROVIDER_ENUM,
+
+    OCO_TOKENS_MAX_INPUT: parseConfigVarValue(process.env.OCO_TOKENS_MAX_INPUT),
+    OCO_TOKENS_MAX_OUTPUT: parseConfigVarValue(
+      process.env.OCO_TOKENS_MAX_OUTPUT
+    ),
+
+    OCO_DESCRIPTION: parseConfigVarValue(process.env.OCO_DESCRIPTION),
+    OCO_EMOJI: parseConfigVarValue(process.env.OCO_EMOJI),
+    OCO_LANGUAGE: process.env.OCO_LANGUAGE,
+    OCO_MESSAGE_TEMPLATE_PLACEHOLDER:
+      process.env.OCO_MESSAGE_TEMPLATE_PLACEHOLDER,
+    OCO_PROMPT_MODULE: process.env.OCO_PROMPT_MODULE as OCO_PROMPT_MODULE_ENUM,
+    OCO_ONE_LINE_COMMIT: parseConfigVarValue(process.env.OCO_ONE_LINE_COMMIT),
+    OCO_TEST_MOCK_TYPE: process.env.OCO_TEST_MOCK_TYPE,
+
+    OCO_GITPUSH: parseConfigVarValue(process.env.OCO_GITPUSH) // todo: deprecate
+  };
+};
+
+export const setGlobalConfig = (
+  config: ConfigType,
+  configPath: string = defaultConfigPath
+) => {
+  writeFileSync(configPath, iniStringify(config), 'utf8');
+};
+
+export const getIsGlobalConfigFileExist = (
+  configPath: string = defaultConfigPath
+) => {
+  return existsSync(configPath);
+};
+
+export const getGlobalConfig = (configPath: string = defaultConfigPath) => {
+  let globalConfig: ConfigType;
+
+  const isGlobalConfigFileExist = getIsGlobalConfigFileExist(configPath);
+  if (!isGlobalConfigFileExist) globalConfig = initGlobalConfig(configPath);
+  else {
+    const configFile = readFileSync(configPath, 'utf8');
+    globalConfig = iniParse(configFile) as ConfigType;
+  }
+
+  return globalConfig;
+};
+
+/**
+ * Merges two configs.
+ * Env config takes precedence over global ~/.opencommit config file
+ * @param main - env config
+ * @param fallback - global ~/.opencommit config file
+ * @returns merged config
+ */
+const mergeConfigs = (main: Partial<ConfigType>, fallback: ConfigType) => {
+  const allKeys = new Set([...Object.keys(main), ...Object.keys(fallback)]);
+  return Array.from(allKeys).reduce((acc, key) => {
+    acc[key] = parseConfigVarValue(main[key] ?? fallback[key]);
+    return acc;
+  }, {} as ConfigType);
+};
+
+interface GetConfigOptions {
+  globalPath?: string;
+  envPath?: string;
+  setDefaultValues?: boolean;
+}
+
+const cleanUndefinedValues = (config: ConfigType) => {
+  return Object.fromEntries(
+    Object.entries(config).map(([_, v]) => {
+      try {
+        if (typeof v === 'string') {
+          if (v === 'undefined') return [_, undefined];
+          if (v === 'null') return [_, null];
+
+          const parsedValue = JSON.parse(v);
+          return [_, parsedValue];
+        }
+        return [_, v];
+      } catch (error) {
+        return [_, v];
+      }
+    })
+  );
+};
+
+export const getConfig = ({
+  envPath = defaultEnvPath,
+  globalPath = defaultConfigPath
+}: GetConfigOptions = {}): ConfigType => {
+  const envConfig = getEnvConfig(envPath);
+  const globalConfig = getGlobalConfig(globalPath);
+
+  const config = mergeConfigs(envConfig, globalConfig);
+
+  const cleanConfig = cleanUndefinedValues(config);
+
+  return cleanConfig as ConfigType;
 };
 
 export const setConfig = (
-  keyValues: [key: string, value: string][],
+  keyValues: [key: string, value: string | boolean | number | null][],
-  configPath: string = defaultConfigPath
+  globalConfigPath: string = defaultConfigPath
 ) => {
-  const config = getConfig() || {};
+  const config = getConfig({
+    globalPath: globalConfigPath
+  });
 
-  for (const [configKey, configValue] of keyValues) {
+  const configToSet = {};
-    if (!configValidators.hasOwnProperty(configKey)) {
+
-      throw new Error(`Unsupported config key: ${configKey}`);
+  for (let [key, value] of keyValues) {
+    if (!configValidators.hasOwnProperty(key)) {
+      const supportedKeys = Object.keys(configValidators).join('\n');
+      throw new Error(
+        `Unsupported config key: ${key}. Expected keys are:\n\n${supportedKeys}.\n\nFor more help refer to our docs: https://github.com/di-sukharev/opencommit`
+      );
     }
 
     let parsedConfigValue;
 
     try {
-      parsedConfigValue = JSON.parse(configValue);
+      if (typeof value === 'string') parsedConfigValue = JSON.parse(value);
+      else parsedConfigValue = value;
     } catch (error) {
-      parsedConfigValue = configValue;
+      parsedConfigValue = value;
     }
 
-    const validValue =
+    const validValue = configValidators[key as CONFIG_KEYS](
-      configValidators[configKey as CONFIG_KEYS](parsedConfigValue);
+      parsedConfigValue,
-    config[configKey as CONFIG_KEYS] = validValue;
+      config
+    );
+
+    configToSet[key] = validValue;
   }
 
-  writeFileSync(configPath, iniStringify(config), 'utf8');
+  setGlobalConfig(mergeConfigs(configToSet, config), globalConfigPath);
 
-  outro(`${chalk.green('✔')} Config successfully set`);
+  outro(`${chalk.green('✔')} config successfully set`);
 };
 
 export const configCommand = command(
@@ -494,9 +590,9 @@ export const configCommand = command(
     parameters: ['<mode>', '<key=values...>']
   },
   async (argv) => {
-    intro('opencommit — config');
     try {
       const { mode, keyValues } = argv._;
+      intro(`COMMAND: config ${mode} ${keyValues}`);
 
       if (mode === CONFIG_MODES.get) {
         const config = getConfig() || {};

src/commands/githook.ts

@@ -1,13 +1,11 @@
+import { intro, outro } from '@clack/prompts';
 import chalk from 'chalk';
 import { command } from 'cleye';
 import { existsSync } from 'fs';
 import fs from 'fs/promises';
 import path from 'path';
-
-import { intro, outro } from '@clack/prompts';
-
-import { COMMANDS } from '../CommandsEnum.js';
 import { assertGitRepo, getCoreHooksPath } from '../utils/git.js';
+import { COMMANDS } from './ENUMS';
 
 const HOOK_NAME = 'prepare-commit-msg';
 const DEFAULT_SYMLINK_URL = path.join('.git', 'hooks', HOOK_NAME);
@@ -94,7 +92,7 @@ export const hookCommand = command(
       }
 
       throw new Error(
-        `Unsupported mode: ${mode}. Supported modes are: 'set' or 'unset', do: \`oco hook set\``
+        `Unsupported mode: ${mode}. Supported modes are: 'set' or 'unset'. Run: \`oco hook set\``
       );
     } catch (error) {
       outro(`${chalk.red('✖')} ${error}`);

src/commands/prepare-commit-msg-hook.ts

@@ -39,10 +39,11 @@ export const prepareCommitMessageHook = async (
 
     const config = getConfig();
 
-    if (!config?.OCO_OPENAI_API_KEY && !config?.OCO_ANTHROPIC_API_KEY && !config?.OCO_AZURE_API_KEY) {
+    if (!config.OCO_API_KEY) {
-      throw new Error(
+      outro(
-        'No OPEN_AI_API or OCO_ANTHROPIC_API_KEY or OCO_AZURE_API_KEY exists. Set your key in ~/.opencommit'
+        'No OCO_API_KEY is set. Set your key via `oco config set OCO_API_KEY=<value>. For more info see https://github.com/di-sukharev/opencommit'
       );
+      return;
     }
 
     const spin = spinner();

src/engine/Engine.ts

@@ -1,7 +1,30 @@
-import { ChatCompletionRequestMessage } from 'openai';
+import AnthropicClient from '@anthropic-ai/sdk';
+import { OpenAIClient as AzureOpenAIClient } from '@azure/openai';
+import { GoogleGenerativeAI as GeminiClient } from '@google/generative-ai';
+import { AxiosInstance as RawAxiosClient } from 'axios';
+import { OpenAI as OpenAIClient } from 'openai';
+import { Mistral as MistralClient } from '@mistralai/mistralai';
+
+export interface AiEngineConfig {
+  apiKey: string;
+  model: string;
+  maxTokensOutput: number;
+  maxTokensInput: number;
+  baseURL?: string;
+}
+
+type Client =
+  | OpenAIClient
+  | AzureOpenAIClient
+  | AnthropicClient
+  | RawAxiosClient
+  | GeminiClient
+  | MistralClient;
 
 export interface AiEngine {
+  config: AiEngineConfig;
+  client: Client;
   generateCommitMessage(
-    messages: Array<ChatCompletionRequestMessage>
+    messages: Array<OpenAIClient.Chat.Completions.ChatCompletionMessageParam>
-  ): Promise<string | undefined>;
+  ): Promise<string | null | undefined>;
 }

src/engine/anthropic.ts

@@ -1,104 +1,62 @@
+import AnthropicClient from '@anthropic-ai/sdk';
+import {
+  MessageCreateParamsNonStreaming,
+  MessageParam
+} from '@anthropic-ai/sdk/resources/messages.mjs';
+import { outro } from '@clack/prompts';
 import axios from 'axios';
 import chalk from 'chalk';
-
+import { OpenAI } from 'openai';
-import Anthropic from '@anthropic-ai/sdk';
-import {ChatCompletionRequestMessage} from 'openai'
-import { MessageCreateParamsNonStreaming, MessageParam } from '@anthropic-ai/sdk/resources';
-
-import { intro, outro } from '@clack/prompts';
-
-import {
-  CONFIG_MODES,
-  DEFAULT_TOKEN_LIMITS,
-  getConfig
-} from '../commands/config';
 import { GenerateCommitMessageErrorEnum } from '../generateCommitMessageFromGitDiff';
 import { tokenCount } from '../utils/tokenCount';
-import { AiEngine } from './Engine';
+import { AiEngine, AiEngineConfig } from './Engine';
-import { MODEL_LIST } from '../commands/config';
 
-const config = getConfig();
+interface AnthropicConfig extends AiEngineConfig {}
 
-const MAX_TOKENS_OUTPUT =
+export class AnthropicEngine implements AiEngine {
-  config?.OCO_TOKENS_MAX_OUTPUT ||
+  config: AnthropicConfig;
-  DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_OUTPUT;
+  client: AnthropicClient;
-const MAX_TOKENS_INPUT =
-  config?.OCO_TOKENS_MAX_INPUT || DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_INPUT;
 
-let provider = config?.OCO_AI_PROVIDER;
+  constructor(config) {
-let apiKey = config?.OCO_ANTHROPIC_API_KEY;
+    this.config = config;
-const [command, mode] = process.argv.slice(2);
+    this.client = new AnthropicClient({ apiKey: this.config.apiKey });
-if (
-  provider === 'anthropic' &&
-  !apiKey &&
-  command !== 'config' &&
-  mode !== CONFIG_MODES.set
-) {
-  intro('opencommit');
-
-  outro(
-    'OCO_ANTHROPIC_API_KEY is not set, please run `oco config set OCO_ANTHROPIC_API_KEY=<your token> . If you are using Claude, make sure you add payment details, so API works.`'
-  );
-  outro(
-    'For help look into README https://github.com/di-sukharev/opencommit#setup'
-  );
-
-  process.exit(1);
-}
-
-const MODEL = config?.OCO_MODEL;
-if (provider === 'anthropic' &&
-  typeof MODEL !== 'string' && 
-  command !== 'config' &&
-  mode !== CONFIG_MODES.set) {
-outro(
-  `${chalk.red('✖')} Unsupported model ${MODEL}. The model can be any string, but the current configuration is not supported.`
-);
-process.exit(1);
-}
-
-export class AnthropicAi implements AiEngine {
-  private anthropicAiApiConfiguration = {
-    apiKey: apiKey
-  };
-  private anthropicAI!: Anthropic;
-
-  constructor() {
-    this.anthropicAI = new Anthropic(this.anthropicAiApiConfiguration);
   }
 
   public generateCommitMessage = async (
-    messages: Array<ChatCompletionRequestMessage>
+    messages: Array<OpenAI.Chat.Completions.ChatCompletionMessageParam>
   ): Promise<string | undefined> => {
-
+    const systemMessage = messages.find((msg) => msg.role === 'system')
-    const systemMessage = messages.find(msg => msg.role === 'system')?.content as string;
+      ?.content as string;
-    const restMessages = messages.filter((msg) => msg.role !== 'system') as MessageParam[];
+    const restMessages = messages.filter(
+      (msg) => msg.role !== 'system'
+    ) as MessageParam[];
 
     const params: MessageCreateParamsNonStreaming = {
-      model: MODEL,
+      model: this.config.model,
       system: systemMessage,
       messages: restMessages,
       temperature: 0,
       top_p: 0.1,
-      max_tokens: MAX_TOKENS_OUTPUT
+      max_tokens: this.config.maxTokensOutput
     };
     try {
       const REQUEST_TOKENS = messages
         .map((msg) => tokenCount(msg.content as string) + 4)
         .reduce((a, b) => a + b, 0);
 
-      if (REQUEST_TOKENS > MAX_TOKENS_INPUT - MAX_TOKENS_OUTPUT) {
+      if (
+        REQUEST_TOKENS >
+        this.config.maxTokensInput - this.config.maxTokensOutput
+      ) {
         throw new Error(GenerateCommitMessageErrorEnum.tooMuchTokens);
       }
 
-      const data  = await this.anthropicAI.messages.create(params);
+      const data = await this.client.messages.create(params);
 
       const message = data?.content[0].text;
 
       return message;
     } catch (error) {
-      outro(`${chalk.red('✖')} ${JSON.stringify(params)}`);
-
       const err = error as Error;
       outro(`${chalk.red('✖')} ${err?.message || err}`);
 

src/engine/azure.ts

@@ -1,81 +1,51 @@
+import {
+  AzureKeyCredential,
+  OpenAIClient as AzureOpenAIClient
+} from '@azure/openai';
+import { outro } from '@clack/prompts';
 import axios from 'axios';
 import chalk from 'chalk';
-import { execa } from 'execa';
+import { OpenAI } from 'openai';
-import {
-  ChatCompletionRequestMessage,
-} from 'openai';
-
-import { OpenAIClient, AzureKeyCredential } from '@azure/openai';
-
-import { intro, outro } from '@clack/prompts';
-
-import {
-  CONFIG_MODES,
-  DEFAULT_TOKEN_LIMITS,
-  getConfig
-} from '../commands/config';
 import { GenerateCommitMessageErrorEnum } from '../generateCommitMessageFromGitDiff';
 import { tokenCount } from '../utils/tokenCount';
-import { AiEngine } from './Engine';
+import { AiEngine, AiEngineConfig } from './Engine';
 
-const config = getConfig();
+interface AzureAiEngineConfig extends AiEngineConfig {
-
+  baseURL: string;
-const MAX_TOKENS_OUTPUT =
+  apiKey: string;
-  config?.OCO_TOKENS_MAX_OUTPUT ||
-  DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_OUTPUT;
-const MAX_TOKENS_INPUT =
-  config?.OCO_TOKENS_MAX_INPUT || DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_INPUT;
-let basePath = config?.OCO_OPENAI_BASE_PATH;
-let apiKey = config?.OCO_AZURE_API_KEY;
-let apiEndpoint = config?.OCO_AZURE_ENDPOINT;
-
-const [command, mode] = process.argv.slice(2);
-
-const provider = config?.OCO_AI_PROVIDER;
-
-if (
-  provider === 'azure' &&
-  !apiKey &&
-  !apiEndpoint &&
-  command !== 'config' &&
-  mode !== CONFIG_MODES.set
-) {
-  intro('opencommit');
-
-  outro(
-    'OCO_AZURE_API_KEY or OCO_AZURE_ENDPOINT are not set, please run `oco config set OCO_AZURE_API_KEY=<your token> . If you are using GPT, make sure you add payment details, so API works.`'
-  );
-  outro(
-    'For help look into README https://github.com/di-sukharev/opencommit#setup'
-  );
-
-  process.exit(1);
 }
 
-const MODEL = config?.OCO_MODEL || 'gpt-3.5-turbo';
+export class AzureEngine implements AiEngine {
+  config: AzureAiEngineConfig;
+  client: AzureOpenAIClient;
 
-export class Azure implements AiEngine {
+  constructor(config: AzureAiEngineConfig) {
-  private openAI!: OpenAIClient;
+    this.config = config;
-
+    this.client = new AzureOpenAIClient(
-  constructor() {
+      this.config.baseURL,
-    if (provider === 'azure') {
+      new AzureKeyCredential(this.config.apiKey)
-      this.openAI = new OpenAIClient(apiEndpoint, new AzureKeyCredential(apiKey));
+    );
-    }
   }
 
-  public generateCommitMessage = async (
+  generateCommitMessage = async (
-    messages: Array<ChatCompletionRequestMessage>
+    messages: Array<OpenAI.Chat.Completions.ChatCompletionMessageParam>
   ): Promise<string | undefined> => {
     try {
       const REQUEST_TOKENS = messages
-        .map((msg) => tokenCount(msg.content) + 4)
+        .map((msg) => tokenCount(msg.content as string) + 4)
         .reduce((a, b) => a + b, 0);
 
-      if (REQUEST_TOKENS > MAX_TOKENS_INPUT - MAX_TOKENS_OUTPUT) {
+      if (
+        REQUEST_TOKENS >
+        this.config.maxTokensInput - this.config.maxTokensOutput
+      ) {
         throw new Error(GenerateCommitMessageErrorEnum.tooMuchTokens);
       }
 
-      const data = await this.openAI.getChatCompletions(MODEL, messages);
+      const data = await this.client.getChatCompletions(
+        this.config.model,
+        messages
+      );
 
       const message = data.choices[0].message;
 
@@ -84,10 +54,10 @@ export class Azure implements AiEngine {
       }
       return message?.content;
     } catch (error) {
-      outro(`${chalk.red('✖')} ${MODEL}`);
+      outro(`${chalk.red('✖')} ${this.config.model}`);
 
       const err = error as Error;
-      outro(`${chalk.red('✖')} ${err?.message || err}`);
+      outro(`${chalk.red('✖')} ${JSON.stringify(error)}`);
 
       if (
         axios.isAxiosError<{ error?: { message: string } }>(error) &&
@@ -105,5 +75,3 @@ export class Azure implements AiEngine {
     }
   };
 }
-
-export const azure = new Azure();

src/engine/flowise.ts

@@ -1,38 +1,40 @@
-import axios, { AxiosError } from 'axios';
+import axios, { AxiosInstance } from 'axios';
-import { ChatCompletionRequestMessage } from 'openai';
+import { OpenAI } from 'openai';
-import { AiEngine } from './Engine';
+import { AiEngine, AiEngineConfig } from './Engine';
 
-import {
+interface FlowiseAiConfig extends AiEngineConfig {}
-  getConfig
-} from '../commands/config';
 
-const config = getConfig();
+export class FlowiseEngine implements AiEngine {
+  config: FlowiseAiConfig;
+  client: AxiosInstance;
 
-export class FlowiseAi implements AiEngine {
+  constructor(config) {
+    this.config = config;
+    this.client = axios.create({
+      url: `${config.baseURL}/${config.apiKey}`,
+      headers: { 'Content-Type': 'application/json' }
+    });
+  }
 
   async generateCommitMessage(
-    messages: Array<ChatCompletionRequestMessage>
+    messages: Array<OpenAI.Chat.Completions.ChatCompletionMessageParam>
   ): Promise<string | undefined> {
+    const gitDiff = (messages[messages.length - 1]?.content as string)
+      .replace(/\\/g, '\\\\')
+      .replace(/"/g, '\\"')
+      .replace(/\n/g, '\\n')
+      .replace(/\r/g, '\\r')
+      .replace(/\t/g, '\\t');
 
-    const gitDiff = messages[ messages.length - 1 ]?.content?.replace(/\\/g, '\\\\')  
-                                                            .replace(/"/g, '\\"')    
-                                                            .replace(/\n/g, '\\n')   
-                                                            .replace(/\r/g, '\\r')   
-                                                            .replace(/\t/g, '\\t');  
-    const url = `http://${config?.OCO_FLOWISE_ENDPOINT}/api/v1/prediction/${config?.OCO_FLOWISE_API_KEY}`; 
     const payload = {
-        question : gitDiff,
+      question: gitDiff,
-        overrideConfig : {
+      overrideConfig: {
-          systemMessagePrompt: messages[0]?.content,
+        systemMessagePrompt: messages[0]?.content
-        },
+      },
-        history : messages.slice( 1, -1 ) 
+      history: messages.slice(1, -1)
-    }
+    };
     try {
-      const response = await axios.post(url, payload, {
+      const response = await this.client.post('', payload);
-        headers: {
-          'Content-Type': 'application/json'
-        }
-      });
       const message = response.data;
       return message?.text;
     } catch (err: any) {

src/engine/gemini.ts

@@ -1,133 +1,88 @@
-import { ChatCompletionRequestMessage } from 'openai';
+import {
-import { AiEngine } from './Engine';
+  Content,
-import { Content, GenerativeModel, GoogleGenerativeAI, HarmBlockThreshold, HarmCategory, Part } from '@google/generative-ai';
+  GoogleGenerativeAI,
-import { CONFIG_MODES, ConfigType, DEFAULT_TOKEN_LIMITS, getConfig, MODEL_LIST } from '../commands/config';
+  HarmBlockThreshold,
-import { intro, outro } from '@clack/prompts';
+  HarmCategory,
-import chalk from 'chalk';
+  Part
+} from '@google/generative-ai';
 import axios from 'axios';
+import { OpenAI } from 'openai';
+import { AiEngine, AiEngineConfig } from './Engine';
 
+interface GeminiConfig extends AiEngineConfig {}
 
-export class Gemini implements AiEngine {
+export class GeminiEngine implements AiEngine {
+  config: GeminiConfig;
+  client: GoogleGenerativeAI;
 
-  private readonly config: ConfigType;
+  constructor(config) {
-  private readonly googleGenerativeAi: GoogleGenerativeAI;
+    this.client = new GoogleGenerativeAI(config.apiKey);
-  private ai: GenerativeModel;
+    this.config = config;
-
-  // vars 
-  private maxTokens = {
-    input: DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_INPUT,
-    output: DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_OUTPUT
-  };
-  private basePath: string;
-  private apiKey: string;
-  private model: string;
-
-  constructor() {
-    this.config = getConfig() as ConfigType;
-    this.googleGenerativeAi = new GoogleGenerativeAI(this.config.OCO_GEMINI_API_KEY);
-    
-    this.warmup();
   }
 
-  async generateCommitMessage(messages: ChatCompletionRequestMessage[]): Promise<string | undefined> {
+  async generateCommitMessage(
-    const systemInstruction = messages.filter(m => m.role === 'system')
+    messages: Array<OpenAI.Chat.Completions.ChatCompletionMessageParam>
-      .map(m => m.content)
+  ): Promise<string | undefined> {
+    const systemInstruction = messages
+      .filter((m) => m.role === 'system')
+      .map((m) => m.content)
       .join('\n');
-    
+
-    this.ai = this.googleGenerativeAi.getGenerativeModel({
+    const gemini = this.client.getGenerativeModel({
-      model: this.model,
+      model: this.config.model,
-      systemInstruction,
+      systemInstruction
     });
-    
+
-    const contents = messages.filter(m => m.role !== 'system')
+    const contents = messages
-      .map(m => ({ parts: [{ text: m.content } as Part], role: m.role == 'user' ? m.role : 'model', } as Content));
+      .filter((m) => m.role !== 'system')
-      
+      .map(
+        (m) =>
+          ({
+            parts: [{ text: m.content } as Part],
+            role: m.role === 'user' ? m.role : 'model'
+          } as Content)
+      );
+
     try {
-      const result = await this.ai.generateContent({
+      const result = await gemini.generateContent({
-        contents, 
+        contents,
         safetySettings: [
-          { 
-            category: HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT, 
-            threshold: HarmBlockThreshold.BLOCK_LOW_AND_ABOVE, 
-          },
           {
-            category: HarmCategory.HARM_CATEGORY_HARASSMENT, 
+            category: HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT,
             threshold: HarmBlockThreshold.BLOCK_LOW_AND_ABOVE
           },
           {
-            category: HarmCategory.HARM_CATEGORY_HATE_SPEECH, 
+            category: HarmCategory.HARM_CATEGORY_HARASSMENT,
             threshold: HarmBlockThreshold.BLOCK_LOW_AND_ABOVE
           },
           {
-            category: HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT, 
+            category: HarmCategory.HARM_CATEGORY_HATE_SPEECH,
             threshold: HarmBlockThreshold.BLOCK_LOW_AND_ABOVE
           },
+          {
+            category: HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT,
+            threshold: HarmBlockThreshold.BLOCK_LOW_AND_ABOVE
+          }
         ],
         generationConfig: {
-          maxOutputTokens: this.maxTokens.output,
+          maxOutputTokens: this.config.maxTokensOutput,
           temperature: 0,
-          topP: 0.1,
+          topP: 0.1
-        },
+        }
       });
-      
+
       return result.response.text();
     } catch (error) {
       const err = error as Error;
-      outro(`${chalk.red('✖')} ${err?.message || err}`);
-      
       if (
         axios.isAxiosError<{ error?: { message: string } }>(error) &&
         error.response?.status === 401
       ) {
         const geminiError = error.response.data.error;
-
+        if (geminiError) throw new Error(geminiError?.message);
-        if (geminiError?.message) outro(geminiError.message);
-        outro(
-          'For help look into README https://github.com/di-sukharev/opencommit#setup'
-        );
       }
-      
+
       throw err;
     }
   }
-
+}
-  private warmup(): void {
-    if (this.config.OCO_TOKENS_MAX_INPUT !== undefined) this.maxTokens.input = this.config.OCO_TOKENS_MAX_INPUT;
-    if (this.config.OCO_TOKENS_MAX_OUTPUT !== undefined) this.maxTokens.output = this.config.OCO_TOKENS_MAX_OUTPUT;
-    this.basePath = this.config.OCO_GEMINI_BASE_PATH;
-    this.apiKey = this.config.OCO_GEMINI_API_KEY;
-
-    const [command, mode] = process.argv.slice(2);
-
-    const provider = this.config.OCO_AI_PROVIDER;
-
-    if (provider === 'gemini' && !this.apiKey &&
-      command !== 'config' && mode !== 'set') {
-      intro('opencommit');
-
-      outro('OCO_GEMINI_API_KEY is not set, please run `oco config set OCO_GEMINI_API_KEY=<your token> . If you are using GPT, make sure you add payment details, so API works.');
-
-      outro(
-        'For help look into README https://github.com/di-sukharev/opencommit#setup'
-      );
-
-      process.exit(1);
-    }
-    
-    this.model = this.config.OCO_MODEL || MODEL_LIST.gemini[0];
-    
-    if (provider === 'gemini' &&
-      !MODEL_LIST.gemini.includes(this.model) &&
-      command !== 'config' &&
-      mode !== CONFIG_MODES.set) {
-      outro(
-        `${chalk.red('✖')} Unsupported model ${this.model} for Gemini. Supported models are: ${MODEL_LIST.gemini.join(
-          ', '
-        )}`
-      );
-
-      process.exit(1);
-    }
-  }
-
-}

src/engine/groq.ts

@@ -0,0 +1,10 @@
+import { OpenAiConfig, OpenAiEngine } from './openAi';
+
+interface GroqConfig extends OpenAiConfig {}
+
+export class GroqEngine extends OpenAiEngine {
+  constructor(config: GroqConfig) {
+    config.baseURL = 'https://api.groq.com/openai/v1';
+    super(config);
+  }
+}

src/engine/mistral.ts

@@ -0,0 +1,82 @@
+import axios from 'axios';
+import { Mistral } from '@mistralai/mistralai';
+import { OpenAI } from 'openai';
+import { GenerateCommitMessageErrorEnum } from '../generateCommitMessageFromGitDiff';
+import { tokenCount } from '../utils/tokenCount';
+import { AiEngine, AiEngineConfig } from './Engine';
+import {
+  AssistantMessage as MistralAssistantMessage,
+  SystemMessage as MistralSystemMessage,
+  ToolMessage as MistralToolMessage,
+  UserMessage as MistralUserMessage
+} from '@mistralai/mistralai/models/components';
+
+export interface MistralAiConfig extends AiEngineConfig {}
+export type MistralCompletionMessageParam = Array<
+| (MistralSystemMessage & { role: "system" })
+| (MistralUserMessage & { role: "user" })
+| (MistralAssistantMessage & { role: "assistant" })
+| (MistralToolMessage & { role: "tool" })
+>
+
+export class MistralAiEngine implements AiEngine {
+  config: MistralAiConfig;
+  client: Mistral;
+
+  constructor(config: MistralAiConfig) {
+    this.config = config;
+
+    if (!config.baseURL) {
+      this.client = new Mistral({ apiKey: config.apiKey });
+    } else {
+      this.client = new Mistral({ apiKey: config.apiKey, serverURL: config.baseURL });
+    }
+  }
+
+  public generateCommitMessage = async (
+    messages: Array<OpenAI.Chat.Completions.ChatCompletionMessageParam>
+  ): Promise<string | null> => {
+    const params = {
+      model: this.config.model,
+      messages: messages as MistralCompletionMessageParam,
+      topP: 0.1,
+      maxTokens: this.config.maxTokensOutput
+    };
+
+    try {
+      const REQUEST_TOKENS = messages
+        .map((msg) => tokenCount(msg.content as string) + 4)
+        .reduce((a, b) => a + b, 0);
+
+      if (
+        REQUEST_TOKENS >
+        this.config.maxTokensInput - this.config.maxTokensOutput
+      )
+        throw new Error(GenerateCommitMessageErrorEnum.tooMuchTokens);
+
+      const completion = await this.client.chat.complete(params);
+
+      if (!completion.choices)
+        throw Error('No completion choice available.')
+      
+      const message = completion.choices[0].message;
+
+      if (!message || !message.content)
+        throw Error('No completion choice available.')
+
+      return message.content as string;
+    } catch (error) {
+      const err = error as Error;
+      if (
+        axios.isAxiosError<{ error?: { message: string } }>(error) &&
+        error.response?.status === 401
+      ) {
+        const mistralError = error.response.data.error;
+
+        if (mistralError) throw new Error(mistralError.message);
+      }
+
+      throw err;
+    }
+  };
+}

src/engine/mlx.ts

@@ -0,0 +1,47 @@
+import axios, { AxiosInstance } from 'axios';
+import { OpenAI } from 'openai';
+import { AiEngine, AiEngineConfig } from './Engine';
+import { chown } from 'fs';
+
+interface MLXConfig extends AiEngineConfig {}
+
+export class MLXEngine implements AiEngine {
+    config: MLXConfig;
+    client: AxiosInstance;
+
+    constructor(config) {
+        this.config = config;
+        this.client = axios.create({
+            url: config.baseURL
+            ? `${config.baseURL}/${config.apiKey}`
+            : 'http://localhost:8080/v1/chat/completions',
+            headers: { 'Content-Type': 'application/json' }
+        });
+    }
+
+    async generateCommitMessage(
+        messages: Array<OpenAI.Chat.Completions.ChatCompletionMessageParam>): 
+        Promise<string | undefined> {
+            const params = {
+                messages,
+                temperature: 0, 
+                top_p: 0.1,
+                repetition_penalty: 1.5,
+                stream: false
+            };
+            try {
+                const response = await this.client.post(
+                    this.client.getUri(this.config),
+                    params
+                );
+                
+                const choices = response.data.choices;
+                const message = choices[0].message;
+                
+                return message?.content;
+            } catch (err: any) {
+                const message = err.response?.data?.error ?? err.message;
+                throw new Error(`MLX provider error: ${message}`);
+            }
+        }
+}

src/engine/ollama.ts

@@ -1,52 +1,44 @@
-import axios, { AxiosError } from 'axios';
+import axios, { AxiosInstance } from 'axios';
-import { ChatCompletionRequestMessage } from 'openai';
+import { OpenAI } from 'openai';
-import { AiEngine } from './Engine';
+import { AiEngine, AiEngineConfig } from './Engine';
 
-import {
+interface OllamaConfig extends AiEngineConfig {}
-  getConfig
-} from '../commands/config';
 
-const config = getConfig();
+export class OllamaEngine implements AiEngine {
+  config: OllamaConfig;
+  client: AxiosInstance;
 
-export class OllamaAi implements AiEngine {
+  constructor(config) {
-  private model = "mistral"; // as default model of Ollama
+    this.config = config;
-  private url = "http://localhost:11434/api/chat"; // default URL of Ollama API
+    this.client = axios.create({
-
+      url: config.baseURL
-  setModel(model: string) {
+        ? `${config.baseURL}/${config.apiKey}`
-    this.model = model ?? config?.OCO_MODEL ?? 'mistral';
+        : 'http://localhost:11434/api/chat',
-  }
+      headers: { 'Content-Type': 'application/json' }
-  
+    });
-  setUrl(url: string) {
-    this.url = url ?? config?.OCO_OLLAMA_API_URL ?? 'http://localhost:11434/api/chat';
   }
+
   async generateCommitMessage(
-    messages: Array<ChatCompletionRequestMessage>
+    messages: Array<OpenAI.Chat.Completions.ChatCompletionMessageParam>
   ): Promise<string | undefined> {
-    const model = this.model;
+    const params = {
-
+      model: this.config.model ?? 'mistral',
-    //console.log(messages);
-    //process.exit()
-
-    const url = this.url;
-    const p = {
-      model,
       messages,
       options: { temperature: 0, top_p: 0.1 },
       stream: false
     };
     try {
-      const response = await axios.post(url, p, {
+      const response = await this.client.post(
-        headers: {
+        this.client.getUri(this.config),
-          'Content-Type': 'application/json'
+        params
-        }
+      );
-      });
 
       const message = response.data.message;
 
       return message?.content;
     } catch (err: any) {
       const message = err.response?.data?.error ?? err.message;
-      throw new Error('local model issues. details: ' + message);
+      throw new Error(`Ollama provider error: ${message}`);
     }
   }
 }

src/engine/openAi.ts

@@ -1,127 +1,64 @@
 import axios from 'axios';
-import chalk from 'chalk';
+import { OpenAI } from 'openai';
-import { execa } from 'execa';
-
-import {
-  ChatCompletionRequestMessage,
-  Configuration as OpenAiApiConfiguration,
-  OpenAIApi
-} from 'openai';
-
-import { intro, outro } from '@clack/prompts';
-
-import {
-  CONFIG_MODES,
-  DEFAULT_TOKEN_LIMITS,
-  getConfig
-} from '../commands/config';
 import { GenerateCommitMessageErrorEnum } from '../generateCommitMessageFromGitDiff';
 import { tokenCount } from '../utils/tokenCount';
-import { AiEngine } from './Engine';
+import { AiEngine, AiEngineConfig } from './Engine';
-import { MODEL_LIST } from '../commands/config';
 
-const config = getConfig();
+export interface OpenAiConfig extends AiEngineConfig {}
 
-const MAX_TOKENS_OUTPUT =
+export class OpenAiEngine implements AiEngine {
-  config?.OCO_TOKENS_MAX_OUTPUT ||
+  config: OpenAiConfig;
-  DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_OUTPUT;
+  client: OpenAI;
-const MAX_TOKENS_INPUT =
-  config?.OCO_TOKENS_MAX_INPUT || DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_INPUT;
-let basePath = config?.OCO_OPENAI_BASE_PATH;
-let apiKey = config?.OCO_OPENAI_API_KEY;
 
-const [command, mode] = process.argv.slice(2);
+  constructor(config: OpenAiConfig) {
+    this.config = config;
 
-const provider = config?.OCO_AI_PROVIDER;
+    if (!config.baseURL) {
-
+      this.client = new OpenAI({ apiKey: config.apiKey });
-if (
+    } else {
-  provider === 'openai' &&
+      this.client = new OpenAI({ apiKey: config.apiKey, baseURL: config.baseURL });
-  !apiKey &&
-  command !== 'config' &&
-  mode !== CONFIG_MODES.set
-) {
-  intro('opencommit');
-
-  outro(
-    'OCO_OPENAI_API_KEY is not set, please run `oco config set OCO_OPENAI_API_KEY=<your token> . If you are using GPT, make sure you add payment details, so API works.`'
-  );
-  outro(
-    'For help look into README https://github.com/di-sukharev/opencommit#setup'
-  );
-
-  process.exit(1);
-}
-
-const MODEL = config?.OCO_MODEL || 'gpt-3.5-turbo';
-if (provider === 'openai' &&
-  typeof MODEL !== 'string' && 
-    command !== 'config' &&
-    mode !== CONFIG_MODES.set) {
-  outro(
-    `${chalk.red('✖')} Unsupported model ${MODEL}. The model can be any string, but the current configuration is not supported.`
-  );
-  process.exit(1);
-}
-
-export class OpenAi implements AiEngine {
-  
-  private openAiApiConfiguration = new OpenAiApiConfiguration({
-    apiKey: apiKey
-  });
-  private openAI!: OpenAIApi;
-
-  constructor() {
-    if (basePath) {
-      this.openAiApiConfiguration.basePath = basePath;
     }
-    this.openAI = new OpenAIApi(this.openAiApiConfiguration);
   }
 
   public generateCommitMessage = async (
-    messages: Array<ChatCompletionRequestMessage>
+    messages: Array<OpenAI.Chat.Completions.ChatCompletionMessageParam>
-  ): Promise<string | undefined> => {
+  ): Promise<string | null> => {
     const params = {
-      model: MODEL,
+      model: this.config.model,
       messages,
       temperature: 0,
       top_p: 0.1,
-      max_tokens: MAX_TOKENS_OUTPUT
+      max_tokens: this.config.maxTokensOutput
     };
+
     try {
       const REQUEST_TOKENS = messages
         .map((msg) => tokenCount(msg.content as string) + 4)
         .reduce((a, b) => a + b, 0);
 
-      if (REQUEST_TOKENS > MAX_TOKENS_INPUT - MAX_TOKENS_OUTPUT) {
+      if (
+        REQUEST_TOKENS >
+        this.config.maxTokensInput - this.config.maxTokensOutput
+      )
         throw new Error(GenerateCommitMessageErrorEnum.tooMuchTokens);
-      }
 
-      const { data } = await this.openAI.createChatCompletion(params);
+      const completion = await this.client.chat.completions.create(params);
 
-      const message = data.choices[0].message;
+      const message = completion.choices[0].message;
 
       return message?.content;
     } catch (error) {
-      outro(`${chalk.red('✖')} ${JSON.stringify(params)}`);
-
       const err = error as Error;
-      outro(`${chalk.red('✖')} ${err?.message || err}`);
-
       if (
         axios.isAxiosError<{ error?: { message: string } }>(error) &&
         error.response?.status === 401
       ) {
         const openAiError = error.response.data.error;
 
-        if (openAiError?.message) outro(openAiError.message);
+        if (openAiError) throw new Error(openAiError.message);
-        outro(
-          'For help look into README https://github.com/di-sukharev/opencommit#setup'
-        );
       }
 
       throw err;
     }
   };
-  
 }
-

src/engine/testAi.ts

@@ -1,31 +1,47 @@
-import { ChatCompletionRequestMessage } from 'openai';
+import { OpenAI } from 'openai';
+
 import { AiEngine } from './Engine';
-import { getConfig } from '../commands/config';
 
 export const TEST_MOCK_TYPES = [
   'commit-message',
-  'prompt-module-commitlint-config',
+  'prompt-module-commitlint-config'
-] as const
+] as const;
-type TestMockType = typeof TEST_MOCK_TYPES[number];
+
+export type TestMockType = (typeof TEST_MOCK_TYPES)[number];
+
+type TestAiEngine = Partial<AiEngine> & {
+  mockType: TestMockType;
+};
+
+export class TestAi implements TestAiEngine {
+  mockType: TestMockType;
+
+  // those are not used in the test engine
+  config: any;
+  client: any;
+  // ---
+
+  constructor(mockType: TestMockType) {
+    this.mockType = mockType;
+  }
 
-export class TestAi implements AiEngine {
   async generateCommitMessage(
-    _messages: Array<ChatCompletionRequestMessage>
+    _messages: Array<OpenAI.Chat.Completions.ChatCompletionMessageParam>
   ): Promise<string | undefined> {
-    const config = getConfig();
+    switch (this.mockType) {
-    switch (config?.OCO_TEST_MOCK_TYPE as TestMockType | undefined) {
       case 'commit-message':
         return 'fix(testAi.ts): test commit message';
       case 'prompt-module-commitlint-config':
-        return `{\n` +
+        return (
+          `{\n` +
           `  "localLanguage": "english",\n` +
           `  "commitFix": "fix(server): Change 'port' variable to uppercase 'PORT'",\n` +
           `  "commitFeat": "feat(server): Allow server to listen on a port specified through environment variable",\n` +
           `  "commitDescription": "Change 'port' variable to uppercase 'PORT'. Allow server to listen on a port specified through environment variable."\n` +
           `}`
+        );
       default:
-        throw Error('unsupported test mock type')
+        throw Error('unsupported test mock type');
     }
   }
 }
-

src/generateCommitMessageFromGitDiff.ts

@@ -1,31 +1,25 @@
-import {
+import { OpenAI } from 'openai';
-  ChatCompletionRequestMessage,
-  ChatCompletionRequestMessageRoleEnum
-} from 'openai';
-
 import { DEFAULT_TOKEN_LIMITS, getConfig } from './commands/config';
 import { getMainCommitPrompt } from './prompts';
+import { getEngine } from './utils/engine';
 import { mergeDiffs } from './utils/mergeDiffs';
 import { tokenCount } from './utils/tokenCount';
-import { getEngine } from './utils/engine';
 
 const config = getConfig();
-const MAX_TOKENS_INPUT =
+const MAX_TOKENS_INPUT = config.OCO_TOKENS_MAX_INPUT;
-  config?.OCO_TOKENS_MAX_INPUT || DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_INPUT;
+const MAX_TOKENS_OUTPUT = config.OCO_TOKENS_MAX_OUTPUT;
-const MAX_TOKENS_OUTPUT =
-  config?.OCO_TOKENS_MAX_OUTPUT ||
-  DEFAULT_TOKEN_LIMITS.DEFAULT_MAX_TOKENS_OUTPUT;
 
 const generateCommitMessageChatCompletionPrompt = async (
   diff: string,
-  fullGitMojiSpec: boolean
+  fullGitMojiSpec: boolean,
-): Promise<Array<ChatCompletionRequestMessage>> => {
+  context: string
-  const INIT_MESSAGES_PROMPT = await getMainCommitPrompt(fullGitMojiSpec);
+): Promise<Array<OpenAI.Chat.Completions.ChatCompletionMessageParam>> => {
+  const INIT_MESSAGES_PROMPT = await getMainCommitPrompt(fullGitMojiSpec, context);
 
   const chatContextAsCompletionRequest = [...INIT_MESSAGES_PROMPT];
 
   chatContextAsCompletionRequest.push({
-    role: ChatCompletionRequestMessageRoleEnum.User,
+    role: 'user',
     content: diff
   });
 
@@ -43,10 +37,14 @@ const ADJUSTMENT_FACTOR = 20;
 
 export const generateCommitMessageByDiff = async (
   diff: string,
-  fullGitMojiSpec: boolean
+  fullGitMojiSpec: boolean = false,
+  context: string = ""
 ): Promise<string> => {
   try {
-    const INIT_MESSAGES_PROMPT = await getMainCommitPrompt(fullGitMojiSpec);
+    const INIT_MESSAGES_PROMPT = await getMainCommitPrompt(
+      fullGitMojiSpec,
+      context
+    );
 
     const INIT_MESSAGES_PROMPT_LENGTH = INIT_MESSAGES_PROMPT.map(
       (msg) => tokenCount(msg.content as string) + 4
@@ -76,7 +74,8 @@ export const generateCommitMessageByDiff = async (
 
     const messages = await generateCommitMessageChatCompletionPrompt(
       diff,
-      fullGitMojiSpec
+      fullGitMojiSpec,
+      context,
     );
 
     const engine = getEngine();
@@ -181,7 +180,7 @@ export const getCommitMsgsPromisesFromFileDiffs = async (
   // merge multiple files-diffs into 1 prompt to save tokens
   const mergedFilesDiffs = mergeDiffs(diffByFiles, maxDiffLength);
 
-  const commitMessagePromises = [] as Promise<string | undefined>[];
+  const commitMessagePromises = [] as Promise<string | null | undefined>[];
 
   for (const fileDiff of mergedFilesDiffs) {
     if (tokenCount(fileDiff) >= maxDiffLength) {

src/github-action.ts

@@ -1,11 +1,9 @@
-import { unlinkSync, writeFileSync } from 'fs';
-
 import core from '@actions/core';
 import exec from '@actions/exec';
 import github from '@actions/github';
 import { intro, outro } from '@clack/prompts';
 import { PushEvent } from '@octokit/webhooks-types';
-
+import { unlinkSync, writeFileSync } from 'fs';
 import { generateCommitMessageByDiff } from './generateCommitMessageFromGitDiff';
 import { randomIntFromInterval } from './utils/randomIntFromInterval';
 import { sleep } from './utils/sleep';
@@ -54,7 +52,7 @@ async function improveMessagesInChunks(diffsAndSHAs: DiffAndSHA[]) {
   const chunkSize = diffsAndSHAs!.length % 2 === 0 ? 4 : 3;
   outro(`Improving commit messages in chunks of ${chunkSize}.`);
   const improvePromises = diffsAndSHAs!.map((commit) =>
-    generateCommitMessageByDiff(commit.diff)
+    generateCommitMessageByDiff(commit.diff, false)
   );
 
   let improvedMessagesAndSHAs: MsgAndSHA[] = [];

src/migrations/00_use_single_api_key_and_url.ts

@@ -0,0 +1,45 @@
+import {
+  CONFIG_KEYS,
+  getConfig,
+  OCO_AI_PROVIDER_ENUM,
+  setConfig
+} from '../commands/config';
+
+export default function () {
+  const config = getConfig({ setDefaultValues: false });
+
+  const aiProvider = config.OCO_AI_PROVIDER;
+
+  let apiKey: string | undefined;
+  let apiUrl: string | undefined;
+
+  if (aiProvider === OCO_AI_PROVIDER_ENUM.OLLAMA) {
+    apiKey = config['OCO_OLLAMA_API_KEY'];
+    apiUrl = config['OCO_OLLAMA_API_URL'];
+  } else if (aiProvider === OCO_AI_PROVIDER_ENUM.ANTHROPIC) {
+    apiKey = config['OCO_ANTHROPIC_API_KEY'];
+    apiUrl = config['OCO_ANTHROPIC_BASE_PATH'];
+  } else if (aiProvider === OCO_AI_PROVIDER_ENUM.OPENAI) {
+    apiKey = config['OCO_OPENAI_API_KEY'];
+    apiUrl = config['OCO_OPENAI_BASE_PATH'];
+  } else if (aiProvider === OCO_AI_PROVIDER_ENUM.AZURE) {
+    apiKey = config['OCO_AZURE_API_KEY'];
+    apiUrl = config['OCO_AZURE_ENDPOINT'];
+  } else if (aiProvider === OCO_AI_PROVIDER_ENUM.GEMINI) {
+    apiKey = config['OCO_GEMINI_API_KEY'];
+    apiUrl = config['OCO_GEMINI_BASE_PATH'];
+  } else if (aiProvider === OCO_AI_PROVIDER_ENUM.FLOWISE) {
+    apiKey = config['OCO_FLOWISE_API_KEY'];
+    apiUrl = config['OCO_FLOWISE_ENDPOINT'];
+  } else {
+    throw new Error(
+      `Migration failed, set AI provider first. Run "oco config set OCO_AI_PROVIDER=<provider>", where <provider> is one of: ${Object.values(
+        OCO_AI_PROVIDER_ENUM
+      ).join(', ')}`
+    );
+  }
+
+  if (apiKey) setConfig([[CONFIG_KEYS.OCO_API_KEY, apiKey]]);
+
+  if (apiUrl) setConfig([[CONFIG_KEYS.OCO_API_URL, apiUrl]]);
+}

src/migrations/01_remove_obsolete_config_keys_from_global_file.ts

@@ -0,0 +1,26 @@
+import { getGlobalConfig, setGlobalConfig } from '../commands/config';
+
+export default function () {
+  const obsoleteKeys = [
+    'OCO_OLLAMA_API_KEY',
+    'OCO_OLLAMA_API_URL',
+    'OCO_ANTHROPIC_API_KEY',
+    'OCO_ANTHROPIC_BASE_PATH',
+    'OCO_OPENAI_API_KEY',
+    'OCO_OPENAI_BASE_PATH',
+    'OCO_AZURE_API_KEY',
+    'OCO_AZURE_ENDPOINT',
+    'OCO_GEMINI_API_KEY',
+    'OCO_GEMINI_BASE_PATH',
+    'OCO_FLOWISE_API_KEY',
+    'OCO_FLOWISE_ENDPOINT'
+  ];
+
+  const globalConfig = getGlobalConfig();
+
+  const configToOverride = { ...globalConfig };
+
+  for (const key of obsoleteKeys) delete configToOverride[key];
+
+  setGlobalConfig(configToOverride);
+}

src/migrations/02_set_missing_default_values.ts

@@ -0,0 +1,22 @@
+import {
+  ConfigType,
+  DEFAULT_CONFIG,
+  getGlobalConfig,
+  setConfig
+} from '../commands/config';
+
+export default function () {
+  const setDefaultConfigValues = (config: ConfigType) => {
+    const entriesToSet: [key: string, value: string | boolean | number][] = [];
+    for (const entry of Object.entries(DEFAULT_CONFIG)) {
+      const [key, _value] = entry;
+      if (config[key] === 'undefined' || config[key] === undefined)
+        entriesToSet.push(entry);
+    }
+
+    if (entriesToSet.length > 0) setConfig(entriesToSet);
+    console.log(entriesToSet);
+  };
+
+  setDefaultConfigValues(getGlobalConfig());
+}

src/migrations/_migrations.ts

@@ -0,0 +1,18 @@
+import migration00 from './00_use_single_api_key_and_url';
+import migration01 from './01_remove_obsolete_config_keys_from_global_file';
+import migration02 from './02_set_missing_default_values';
+
+export const migrations = [
+  {
+    name: '00_use_single_api_key_and_url',
+    run: migration00
+  },
+  {
+    name: '01_remove_obsolete_config_keys_from_global_file',
+    run: migration01
+  },
+  {
+    name: '02_set_missing_default_values',
+    run: migration02
+  }
+];

src/migrations/_run.ts

@@ -0,0 +1,71 @@
+import fs from 'fs';
+import { homedir } from 'os';
+import { join as pathJoin } from 'path';
+import { migrations } from './_migrations';
+import { outro } from '@clack/prompts';
+import chalk from 'chalk';
+import {
+  getConfig,
+  getIsGlobalConfigFileExist,
+  OCO_AI_PROVIDER_ENUM
+} from '../commands/config';
+
+const migrationsFile = pathJoin(homedir(), '.opencommit_migrations');
+
+const getCompletedMigrations = (): string[] => {
+  if (!fs.existsSync(migrationsFile)) {
+    return [];
+  }
+  const data = fs.readFileSync(migrationsFile, 'utf-8');
+  return data ? JSON.parse(data) : [];
+};
+
+const saveCompletedMigration = (migrationName: string) => {
+  const completedMigrations = getCompletedMigrations();
+  completedMigrations.push(migrationName);
+  fs.writeFileSync(
+    migrationsFile,
+    JSON.stringify(completedMigrations, null, 2)
+  );
+};
+
+export const runMigrations = async () => {
+  // if no config file, we assume it's a new installation and no migrations are needed
+  if (!getIsGlobalConfigFileExist()) return;
+
+  const config = getConfig();
+  if (config.OCO_AI_PROVIDER === OCO_AI_PROVIDER_ENUM.TEST) return;
+
+  const completedMigrations = getCompletedMigrations();
+
+  let isMigrated = false;
+
+  for (const migration of migrations) {
+    if (!completedMigrations.includes(migration.name)) {
+      try {
+        console.log('Applying migration', migration.name);
+        migration.run();
+        console.log('Migration applied successfully', migration.name);
+        saveCompletedMigration(migration.name);
+      } catch (error) {
+        outro(
+          `${chalk.red('Failed to apply migration')} ${
+            migration.name
+          }: ${error}`
+        );
+        process.exit(1);
+      }
+
+      isMigrated = true;
+    }
+  }
+
+  if (isMigrated) {
+    outro(
+      `${chalk.green(
+        '✔'
+      )} Migrations to your config were applied successfully. Please rerun.`
+    );
+    process.exit(0);
+  }
+};

src/modules/commitlint/config.ts

@@ -2,16 +2,16 @@ import { spinner } from '@clack/prompts';
 
 import { getConfig } from '../../commands/config';
 import { i18n, I18nLocals } from '../../i18n';
+import { getEngine } from '../../utils/engine';
 import { COMMITLINT_LLM_CONFIG_PATH } from './constants';
 import { computeHash } from './crypto';
 import { commitlintPrompts, inferPromptsFromCommitlintConfig } from './prompts';
 import { getCommitLintPWDConfig } from './pwd-commitlint';
 import { CommitlintLLMConfig } from './types';
 import * as utils from './utils';
-import { getEngine } from '../../utils/engine';
 
 const config = getConfig();
-const translation = i18n[(config?.OCO_LANGUAGE as I18nLocals) || 'en'];
+const translation = i18n[(config.OCO_LANGUAGE as I18nLocals) || 'en'];
 
 export const configureCommitlintIntegration = async (force = false) => {
   const spin = spinner();
@@ -26,7 +26,7 @@ export const configureCommitlintIntegration = async (force = false) => {
       * @commitlint >= 9.0.0 is installed in the local directory.
       * 'node_modules/@commitlint/load' package exists.
       * A valid @commitlint configuration exists.
-      `,
+      `
     );
   }
 

src/modules/commitlint/prompts.ts

@@ -1,8 +1,5 @@
 import chalk from 'chalk';
-import {
+import { OpenAI } from 'openai';
-  ChatCompletionRequestMessage,
-  ChatCompletionRequestMessageRoleEnum
-} from 'openai';
 
 import { outro } from '@clack/prompts';
 import {
@@ -17,7 +14,7 @@ import { i18n, I18nLocals } from '../../i18n';
 import { IDENTITY, INIT_DIFF_PROMPT } from '../../prompts';
 
 const config = getConfig();
-const translation = i18n[(config?.OCO_LANGUAGE as I18nLocals) || 'en'];
+const translation = i18n[(config.OCO_LANGUAGE as I18nLocals) || 'en'];
 
 type DeepPartial<T> = {
   [P in keyof T]?: {
@@ -214,11 +211,10 @@ const STRUCTURE_OF_COMMIT = `
 // Prompt to generate LLM-readable rules based on @commitlint rules.
 const GEN_COMMITLINT_CONSISTENCY_PROMPT = (
   prompts: string[]
-): ChatCompletionRequestMessage[] => [
+): OpenAI.Chat.Completions.ChatCompletionMessageParam[] => [
   {
-    role: ChatCompletionRequestMessageRoleEnum.Assistant,
+    role: 'system',
-    // prettier-ignore
+    content: `${IDENTITY} Your mission is to create clean and comprehensive commit messages for two different changes in a single codebase and output them in the provided JSON format: one for a bug fix and another for a new feature.
-    content: `${IDENTITY} Your mission is to create clean and comprehensive commit messages for two different changes in a single codebase and output them in the provided JSON format: one for a bug fix and another for a new feature. 
 
 Here are the specific requirements and conventions that should be strictly followed:
 
@@ -260,22 +256,33 @@ Example Git Diff is to follow:`
 const INIT_MAIN_PROMPT = (
   language: string,
   prompts: string[]
-): ChatCompletionRequestMessage => ({
+): OpenAI.Chat.Completions.ChatCompletionMessageParam => ({
-  role: ChatCompletionRequestMessageRoleEnum.System,
+  role: 'system',
-  // prettier-ignore
+  content: `${IDENTITY} Your mission is to create clean and comprehensive commit messages in the given @commitlint convention and explain WHAT were the changes ${
-  content: `${IDENTITY} Your mission is to create clean and comprehensive commit messages in the given @commitlint convention and explain WHAT were the changes and WHY the changes were done. I'll send you an output of 'git diff --staged' command, and you convert it into a commit message.
+    config.OCO_WHY ? 'and WHY the changes were done' : ''
-${config?.OCO_EMOJI ? 'Use GitMoji convention to preface the commit.' : 'Do not preface the commit with anything.'}
+  }. I'll send you an output of 'git diff --staged' command, and you convert it into a commit message.
-${config?.OCO_DESCRIPTION ? 'Add a short description of WHY the changes are done after the commit message. Don\'t start it with "This commit", just describe the changes.' : "Don't add any descriptions to the commit, only commit message."}
+${
+  config.OCO_EMOJI
+    ? 'Use GitMoji convention to preface the commit.'
+    : 'Do not preface the commit with anything.'
+}
+${
+  config.OCO_DESCRIPTION
+    ? 'Add a short description of WHY the changes are done after the commit message. Don\'t start it with "This commit", just describe the changes.'
+    : "Don't add any descriptions to the commit, only commit message."
+}
 Use the present tense. Use ${language} to answer.
-${ config?.OCO_ONE_LINE_COMMIT ? 'Craft a concise commit message that encapsulates all changes made, with an emphasis on the primary updates. If the modifications share a common theme or scope, mention it succinctly; otherwise, leave the scope out to maintain focus. The goal is to provide a clear and unified overview of the changes in a one single message, without diverging into a list of commit per file change.' : ""}
+${
+  config.OCO_ONE_LINE_COMMIT
+    ? 'Craft a concise commit message that encapsulates all changes made, with an emphasis on the primary updates. If the modifications share a common theme or scope, mention it succinctly; otherwise, leave the scope out to maintain focus. The goal is to provide a clear and unified overview of the changes in a one single message, without diverging into a list of commit per file change.'
+    : ''
+}
     
 You will strictly follow the following conventions to generate the content of the commit message:
 - ${prompts.join('\n- ')}
 
 The conventions refers to the following structure of commit message:
-${STRUCTURE_OF_COMMIT}
+${STRUCTURE_OF_COMMIT}`
-    
-    `
 });
 
 export const commitlintPrompts = {

src/modules/commitlint/pwd-commitlint.ts

@@ -1,13 +1,29 @@
 import fs from 'fs/promises';
 import path from 'path';
 
+const findModulePath = (moduleName: string) => {
+  const searchPaths = [
+    path.join('node_modules', moduleName),
+    path.join('node_modules', '.pnpm')
+  ];
+
+  for (const basePath of searchPaths) {
+    try {
+      const resolvedPath = require.resolve(moduleName, { paths: [basePath] });
+      return resolvedPath;
+    } catch {
+      // Continue to the next search path if the module is not found
+    }
+  }
+
+  throw new Error(`Cannot find module ${moduleName}`);
+};
+
 const getCommitLintModuleType = async (): Promise<'cjs' | 'esm'> => {
-  const packageFile = 'node_modules/@commitlint/load/package.json';
+  const packageFile = '@commitlint/load/package.json';
-  const packageJsonPath = path.join(
+  const packageJsonPath = findModulePath(packageFile);
-    process.env.PWD || process.cwd(),
-    packageFile,
-  );
   const packageJson = JSON.parse(await fs.readFile(packageJsonPath, 'utf8'));
+
   if (!packageJson) {
     throw new Error(`Failed to parse ${packageFile}`);
   }
@@ -19,7 +35,7 @@ const getCommitLintModuleType = async (): Promise<'cjs' | 'esm'> => {
  * QualifiedConfig from any version of @commitlint/types
  * @see https://github.com/conventional-changelog/commitlint/blob/master/@commitlint/types/src/load.ts
  */
-type QualifiedConfigOnAnyVersion = { [key:string]: unknown };
+type QualifiedConfigOnAnyVersion = { [key: string]: unknown };
 
 /**
  *  This code is loading the configuration for the `@commitlint` package from the current working
@@ -27,36 +43,31 @@ type QualifiedConfigOnAnyVersion = { [key:string]: unknown };
  *
  * @returns
  */
-export const getCommitLintPWDConfig = async (): Promise<QualifiedConfigOnAnyVersion | null> => {
+export const getCommitLintPWDConfig =
-  let load, nodeModulesPath;
+  async (): Promise<QualifiedConfigOnAnyVersion | null> => {
-  switch (await getCommitLintModuleType()) {
+    let load: Function, modulePath: string;
-    case 'cjs':
+    switch (await getCommitLintModuleType()) {
-      /**
+      case 'cjs':
-       * CommonJS (<= commitlint@v18.x.x.)
+        /**
-       */
+         * CommonJS (<= commitlint@v18.x.x.)
-      nodeModulesPath = path.join(
+         */
-        process.env.PWD || process.cwd(),
+        modulePath = findModulePath('@commitlint/load');
-        'node_modules/@commitlint/load',
+        load = require(modulePath).default;
-      );
+        break;
-      load = require(nodeModulesPath).default;
+      case 'esm':
-      break;
+        /**
-    case 'esm':
+         * ES Module (commitlint@v19.x.x. <= )
-      /**
+         * Directory import is not supported in ES Module resolution, so import the file directly
-       * ES Module (commitlint@v19.x.x. <= )
+         */
-       * Directory import is not supported in ES Module resolution, so import the file directly
+        modulePath = await findModulePath('@commitlint/load/lib/load.js');
-       */
+        load = (await import(modulePath)).default;
-      nodeModulesPath = path.join(
+        break;
-        process.env.PWD || process.cwd(),
+    }
-        'node_modules/@commitlint/load/lib/load.js',
-      );
-      load = (await import(nodeModulesPath)).default;
-      break;
-  }
 
-  if (load && typeof load === 'function') {
+    if (load && typeof load === 'function') {
-    return await load();
+      return await load();
-  }
+    }
 
-  // @commitlint/load is not a function
+    // @commitlint/load is not a function
-  return null;
+    return null;
-};
+  };

src/prompts.ts

@@ -1,10 +1,5 @@
-import {
-  ChatCompletionRequestMessage,
-  ChatCompletionRequestMessageRoleEnum
-} from 'openai';
-
 import { note } from '@clack/prompts';
-
+import { OpenAI } from 'openai';
 import { getConfig } from './commands/config';
 import { i18n, I18nLocals } from './i18n';
 import { configureCommitlintIntegration } from './modules/commitlint/config';
@@ -14,117 +9,148 @@ import * as utils from './modules/commitlint/utils';
 import { removeConventionalCommitWord } from './utils/removeConventionalCommitWord';
 
 const config = getConfig();
-const translation = i18n[(config?.OCO_LANGUAGE as I18nLocals) || 'en'];
+const translation = i18n[(config.OCO_LANGUAGE as I18nLocals) || 'en'];
 
 export const IDENTITY =
-  'You are to act as the author of a commit message in git.';
+  'You are to act as an author of a commit message in git.';
+
+const GITMOJI_HELP = `Use GitMoji convention to preface the commit. Here are some help to choose the right emoji (emoji, description): 
+🐛, Fix a bug; 
+✨, Introduce new features; 
+📝, Add or update documentation; 
+🚀, Deploy stuff; 
+✅, Add, update, or pass tests; 
+♻️, Refactor code; 
+⬆️, Upgrade dependencies; 
+🔧, Add or update configuration files; 
+🌐, Internationalization and localization; 
+💡, Add or update comments in source code;`;
+
+const FULL_GITMOJI_SPEC = `${GITMOJI_HELP}
+🎨, Improve structure / format of the code; 
+⚡️, Improve performance; 
+🔥, Remove code or files; 
+🚑️, Critical hotfix; 
+💄, Add or update the UI and style files; 
+🎉, Begin a project; 
+🔒️, Fix security issues; 
+🔐, Add or update secrets; 
+🔖, Release / Version tags; 
+🚨, Fix compiler / linter warnings; 
+🚧, Work in progress; 
+💚, Fix CI Build; 
+⬇️, Downgrade dependencies; 
+📌, Pin dependencies to specific versions; 
+👷, Add or update CI build system; 
+📈, Add or update analytics or track code; 
+➕, Add a dependency; 
+➖, Remove a dependency; 
+🔨, Add or update development scripts; 
+✏️, Fix typos; 
+💩, Write bad code that needs to be improved; 
+⏪️, Revert changes; 
+🔀, Merge branches; 
+📦️, Add or update compiled files or packages; 
+👽️, Update code due to external API changes; 
+🚚, Move or rename resources (e.g.: files, paths, routes); 
+📄, Add or update license; 
+💥, Introduce breaking changes; 
+🍱, Add or update assets; 
+♿️, Improve accessibility; 
+🍻, Write code drunkenly; 
+💬, Add or update text and literals; 
+🗃️, Perform database related changes; 
+🔊, Add or update logs; 
+🔇, Remove logs; 
+👥, Add or update contributor(s); 
+🚸, Improve user experience / usability; 
+🏗️, Make architectural changes; 
+📱, Work on responsive design; 
+🤡, Mock things; 
+🥚, Add or update an easter egg; 
+🙈, Add or update a .gitignore file; 
+📸, Add or update snapshots; 
+⚗️, Perform experiments; 
+🔍️, Improve SEO; 
+🏷️, Add or update types; 
+🌱, Add or update seed files; 
+🚩, Add, update, or remove feature flags; 
+🥅, Catch errors; 
+💫, Add or update animations and transitions; 
+🗑️, Deprecate code that needs to be cleaned up; 
+🛂, Work on code related to authorization, roles and permissions; 
+🩹, Simple fix for a non-critical issue; 
+🧐, Data exploration/inspection; 
+⚰️, Remove dead code; 
+🧪, Add a failing test; 
+👔, Add or update business logic; 
+🩺, Add or update healthcheck; 
+🧱, Infrastructure related changes; 
+🧑‍💻, Improve developer experience; 
+💸, Add sponsorships or money related infrastructure; 
+🧵, Add or update code related to multithreading or concurrency; 
+🦺, Add or update code related to validation.`;
+
+const CONVENTIONAL_COMMIT_KEYWORDS =
+  'Do not preface the commit with anything, except for the conventional commit keywords: fix, feat, build, chore, ci, docs, style, refactor, perf, test.';
+
+const getCommitConvention = (fullGitMojiSpec: boolean) =>
+  config.OCO_EMOJI
+    ? fullGitMojiSpec
+      ? FULL_GITMOJI_SPEC
+      : GITMOJI_HELP
+    : CONVENTIONAL_COMMIT_KEYWORDS;
+
+const getDescriptionInstruction = () =>
+  config.OCO_DESCRIPTION
+    ? 'Add a short description of WHY the changes are done after the commit message. Don\'t start it with "This commit", just describe the changes.'
+    : "Don't add any descriptions to the commit, only commit message.";
+
+const getOneLineCommitInstruction = () =>
+  config.OCO_ONE_LINE_COMMIT
+    ? 'Craft a concise commit message that encapsulates all changes made, with an emphasis on the primary updates. If the modifications share a common theme or scope, mention it succinctly; otherwise, leave the scope out to maintain focus. The goal is to provide a clear and unified overview of the changes in a one single message, without diverging into a list of commit per file change.'
+    : '';
+
+/**
+ * Get the context of the user input
+ * @param extraArgs - The arguments passed to the command line
+ * @example
+ *  $ oco -- This is a context used to generate the commit message
+ * @returns - The context of the user input
+ */
+const userInputCodeContext = (context: string) => {
+  if (context !== '' && context !== ' ') {
+    return `Additional context provided by the user: <context>${context}</context>\nConsider this context when generating the commit message, incorporating relevant information when appropriate.`;
+  }
+  return '';
+};
 
 const INIT_MAIN_PROMPT = (
   language: string,
-  fullGitMojiSpec: boolean
+  fullGitMojiSpec: boolean,
-): ChatCompletionRequestMessage => ({
+  context: string
-  role: ChatCompletionRequestMessageRoleEnum.System,
+): OpenAI.Chat.Completions.ChatCompletionMessageParam => ({
-  content: `${IDENTITY} Your mission is to create clean and comprehensive commit messages as per the ${
+  role: 'system',
-    fullGitMojiSpec ? 'GitMoji specification' : 'conventional commit convention'
+  content: (() => {
-  } and explain WHAT were the changes and mainly WHY the changes were done. I'll send you an output of 'git diff --staged' command, and you are to convert it into a commit message.
+    const commitConvention = fullGitMojiSpec
-  ${
+      ? 'GitMoji specification'
-    config?.OCO_EMOJI
+      : 'Conventional Commit Convention';
-      ? 'Use GitMoji convention to preface the commit. Here are some help to choose the right emoji (emoji, description): ' +
+    const missionStatement = `${IDENTITY} Your mission is to create clean and comprehensive commit messages as per the ${commitConvention} and explain WHAT were the changes and mainly WHY the changes were done.`;
-        '🐛, Fix a bug; ' +
+    const diffInstruction =
-        '✨, Introduce new features; ' +
+      "I'll send you an output of 'git diff --staged' command, and you are to convert it into a commit message.";
-        '📝, Add or update documentation; ' +
+    const conventionGuidelines = getCommitConvention(fullGitMojiSpec);
-        '🚀, Deploy stuff; ' +
+    const descriptionGuideline = getDescriptionInstruction();
-        '✅, Add, update, or pass tests; ' +
+    const oneLineCommitGuideline = getOneLineCommitInstruction();
-        '♻️, Refactor code; ' +
+    const generalGuidelines = `Use the present tense. Lines must not be longer than 74 characters. Use ${language} for the commit message.`;
-        '⬆️, Upgrade dependencies; ' +
+    const userInputContext = userInputCodeContext(context);
-        '🔧, Add or update configuration files; ' +
+
-        '🌐, Internationalization and localization; ' +
+    return `${missionStatement}\n${diffInstruction}\n${conventionGuidelines}\n${descriptionGuideline}\n${oneLineCommitGuideline}\n${generalGuidelines}\n${userInputContext}`;
-        '💡, Add or update comments in source code; ' +
+  })()
-        `${
-          fullGitMojiSpec
-            ? '🎨, Improve structure / format of the code; ' +
-              '⚡️, Improve performance; ' +
-              '🔥, Remove code or files; ' +
-              '🚑️, Critical hotfix; ' +
-              '💄, Add or update the UI and style files; ' +
-              '🎉, Begin a project; ' +
-              '🔒️, Fix security issues; ' +
-              '🔐, Add or update secrets; ' +
-              '🔖, Release / Version tags; ' +
-              '🚨, Fix compiler / linter warnings; ' +
-              '🚧, Work in progress; ' +
-              '💚, Fix CI Build; ' +
-              '⬇️, Downgrade dependencies; ' +
-              '📌, Pin dependencies to specific versions; ' +
-              '👷, Add or update CI build system; ' +
-              '📈, Add or update analytics or track code; ' +
-              '➕, Add a dependency; ' +
-              '➖, Remove a dependency; ' +
-              '🔨, Add or update development scripts; ' +
-              '✏️, Fix typos; ' +
-              '💩, Write bad code that needs to be improved; ' +
-              '⏪️, Revert changes; ' +
-              '🔀, Merge branches; ' +
-              '📦️, Add or update compiled files or packages; ' +
-              '👽️, Update code due to external API changes; ' +
-              '🚚, Move or rename resources (e.g.: files, paths, routes); ' +
-              '📄, Add or update license; ' +
-              '💥, Introduce breaking changes; ' +
-              '🍱, Add or update assets; ' +
-              '♿️, Improve accessibility; ' +
-              '🍻, Write code drunkenly; ' +
-              '💬, Add or update text and literals; ' +
-              '🗃️, Perform database related changes; ' +
-              '🔊, Add or update logs; ' +
-              '🔇, Remove logs; ' +
-              '👥, Add or update contributor(s); ' +
-              '🚸, Improve user experience / usability; ' +
-              '🏗️, Make architectural changes; ' +
-              '📱, Work on responsive design; ' +
-              '🤡, Mock things; ' +
-              '🥚, Add or update an easter egg; ' +
-              '🙈, Add or update a .gitignore file; ' +
-              '📸, Add or update snapshots; ' +
-              '⚗️, Perform experiments; ' +
-              '🔍️, Improve SEO; ' +
-              '🏷️, Add or update types; ' +
-              '🌱, Add or update seed files; ' +
-              '🚩, Add, update, or remove feature flags; ' +
-              '🥅, Catch errors; ' +
-              '💫, Add or update animations and transitions; ' +
-              '🗑️, Deprecate code that needs to be cleaned up; ' +
-              '🛂, Work on code related to authorization, roles and permissions; ' +
-              '🩹, Simple fix for a non-critical issue; ' +
-              '🧐, Data exploration/inspection; ' +
-              '⚰️, Remove dead code; ' +
-              '🧪, Add a failing test; ' +
-              '👔, Add or update business logic; ' +
-              '🩺, Add or update healthcheck; ' +
-              '🧱, Infrastructure related changes; ' +
-              '🧑‍💻, Improve developer experience; ' +
-              '💸, Add sponsorships or money related infrastructure; ' +
-              '🧵, Add or update code related to multithreading or concurrency; ' +
-              '🦺, Add or update code related to validation.'
-            : ''
-        }`
-      : 'Do not preface the commit with anything. Conventional commit keywords:' +
-        'fix, feat, build, chore, ci, docs, style, refactor, perf, test.'
-  }  
-    ${
-      config?.OCO_DESCRIPTION
-        ? 'Add a short description of WHY the changes are done after the commit message. Don\'t start it with "This commit", just describe the changes.'
-        : "Don't add any descriptions to the commit, only commit message."
-    }
-    ${
-      config?.OCO_ONE_LINE_COMMIT
-        ? 'Craft a concise commit message that encapsulates all changes made, with an emphasis on the primary updates. If the modifications share a common theme or scope, mention it succinctly; otherwise, leave the scope out to maintain focus. The goal is to provide a clear and unified overview of the changes in a one single message, without diverging into a list of commit per file change.'
-        : ''
-    }
-    Use the present tense. Lines must not be longer than 74 characters. Use ${language} for the commit message.`
 });
 
-export const INIT_DIFF_PROMPT: ChatCompletionRequestMessage = {
+export const INIT_DIFF_PROMPT: OpenAI.Chat.Completions.ChatCompletionMessageParam =
-  role: ChatCompletionRequestMessageRoleEnum.User,
+{
+  role: 'user',
   content: `diff --git a/src/server.ts b/src/server.ts
     index ad4db42..f3b18a9 100644
     --- a/src/server.ts
@@ -151,27 +177,34 @@ export const INIT_DIFF_PROMPT: ChatCompletionRequestMessage = {
                 });`
 };
 
+const getContent = (translation: ConsistencyPrompt) => {
+  const fix = config.OCO_EMOJI
+    ? `🐛 ${removeConventionalCommitWord(translation.commitFix)}`
+    : translation.commitFix;
+
+  const feat = config.OCO_EMOJI
+    ? `✨ ${removeConventionalCommitWord(translation.commitFeat)}`
+    : translation.commitFeat;
+
+  const description = config.OCO_DESCRIPTION
+    ? translation.commitDescription
+    : '';
+
+  return `${fix}\n${feat}\n${description}`;
+};
+
 const INIT_CONSISTENCY_PROMPT = (
   translation: ConsistencyPrompt
-): ChatCompletionRequestMessage => ({
+): OpenAI.Chat.Completions.ChatCompletionMessageParam => ({
-  role: ChatCompletionRequestMessageRoleEnum.Assistant,
+  role: 'assistant',
-  content: `${
+  content: getContent(translation)
-    config?.OCO_EMOJI
-      ? `🐛 ${removeConventionalCommitWord(translation.commitFix)}`
-      : translation.commitFix
-  }
-${
-  config?.OCO_EMOJI
-    ? `✨ ${removeConventionalCommitWord(translation.commitFeat)}`
-    : translation.commitFeat
-}
-${config?.OCO_DESCRIPTION ? translation.commitDescription : ''}`
 });
 
 export const getMainCommitPrompt = async (
-  fullGitMojiSpec: boolean
+  fullGitMojiSpec: boolean,
-): Promise<ChatCompletionRequestMessage[]> => {
+  context: string
-  switch (config?.OCO_PROMPT_MODULE) {
+): Promise<Array<OpenAI.Chat.Completions.ChatCompletionMessageParam>> => {
+  switch (config.OCO_PROMPT_MODULE) {
     case '@commitlint':
       if (!(await utils.commitlintLLMConfigExists())) {
         note(
@@ -191,15 +224,14 @@ export const getMainCommitPrompt = async (
         INIT_DIFF_PROMPT,
         INIT_CONSISTENCY_PROMPT(
           commitLintConfig.consistency[
-            translation.localLanguage
+          translation.localLanguage
           ] as ConsistencyPrompt
         )
       ];
 
     default:
-      // conventional-commit
       return [
-        INIT_MAIN_PROMPT(translation.localLanguage, fullGitMojiSpec),
+        INIT_MAIN_PROMPT(translation.localLanguage, fullGitMojiSpec, context),
         INIT_DIFF_PROMPT,
         INIT_CONSISTENCY_PROMPT(translation)
       ];

src/utils/engine.ts

@@ -1,37 +1,57 @@
+import { getConfig, OCO_AI_PROVIDER_ENUM } from '../commands/config';
+import { AnthropicEngine } from '../engine/anthropic';
+import { AzureEngine } from '../engine/azure';
 import { AiEngine } from '../engine/Engine';
-import { OpenAi } from '../engine/openAi';
+import { FlowiseEngine } from '../engine/flowise';
-import { Gemini } from '../engine/gemini';
+import { GeminiEngine } from '../engine/gemini';
-import { getConfig } from '../commands/config';
+import { OllamaEngine } from '../engine/ollama';
-import { OllamaAi } from '../engine/ollama';
+import { OpenAiEngine } from '../engine/openAi';
-import { AnthropicAi } from '../engine/anthropic'
+import { MistralAiEngine } from '../engine/mistral';
-import { TestAi } from '../engine/testAi';
+import { TestAi, TestMockType } from '../engine/testAi';
-import { Azure } from '../engine/azure';
+import { GroqEngine } from '../engine/groq';
-import { FlowiseAi } from '../engine/flowise'
+import { MLXEngine } from '../engine/mlx';
 
 export function getEngine(): AiEngine {
   const config = getConfig();
-  const provider = config?.OCO_AI_PROVIDER;
+  const provider = config.OCO_AI_PROVIDER;
-  
+
-  if (provider?.startsWith('ollama')) {
+  const DEFAULT_CONFIG = {
-    const ollamaAi = new OllamaAi();
+    model: config.OCO_MODEL!,
-    const model = provider.substring('ollama/'.length);
+    maxTokensOutput: config.OCO_TOKENS_MAX_OUTPUT!,
-    if (model) {
+    maxTokensInput: config.OCO_TOKENS_MAX_INPUT!,
-      ollamaAi.setModel(model);
+    baseURL: config.OCO_API_URL!,
-      ollamaAi.setUrl(config?.OCO_OLLAMA_API_URL);
+    apiKey: config.OCO_API_KEY!
-    }
+  };
-    return ollamaAi;
+
-  } else if (provider == 'anthropic') {
+  switch (provider) {
-    return new AnthropicAi();
+    case OCO_AI_PROVIDER_ENUM.OLLAMA:
-  } else if (provider == 'test') {
+      return new OllamaEngine(DEFAULT_CONFIG);
-    return new TestAi();
+
-  } else if (provider == 'gemini') {
+    case OCO_AI_PROVIDER_ENUM.ANTHROPIC:
-    return new Gemini();  
+      return new AnthropicEngine(DEFAULT_CONFIG);
-  } else if (provider == 'azure') {
+
-  	return new Azure();
+    case OCO_AI_PROVIDER_ENUM.TEST:
-  } else if( provider == 'flowise'){
+      return new TestAi(config.OCO_TEST_MOCK_TYPE as TestMockType);
-    return new FlowiseAi();
+
+    case OCO_AI_PROVIDER_ENUM.GEMINI:
+      return new GeminiEngine(DEFAULT_CONFIG);
+
+    case OCO_AI_PROVIDER_ENUM.AZURE:
+      return new AzureEngine(DEFAULT_CONFIG);
+
+    case OCO_AI_PROVIDER_ENUM.FLOWISE:
+      return new FlowiseEngine(DEFAULT_CONFIG);
+
+    case OCO_AI_PROVIDER_ENUM.GROQ:
+      return new GroqEngine(DEFAULT_CONFIG);
+
+    case OCO_AI_PROVIDER_ENUM.MISTRAL:
+      return new MistralAiEngine(DEFAULT_CONFIG);
+
+    case OCO_AI_PROVIDER_ENUM.MLX:
+      return new MLXEngine(DEFAULT_CONFIG);
+
+    default:
+      return new OpenAiEngine(DEFAULT_CONFIG);
   }
-  
-  //open ai gpt by default
-  return new OpenAi();
 }

test/e2e/gitPush.test.ts

@@ -0,0 +1,205 @@
+import path from 'path';
+import 'cli-testing-library/extend-expect';
+import { exec } from 'child_process';
+import { prepareTempDir } from './utils';
+import { promisify } from 'util';
+import { render } from 'cli-testing-library';
+import { resolve } from 'path';
+import { rm } from 'fs';
+const fsExec = promisify(exec);
+const fsRemove = promisify(rm);
+
+/**
+ * git remote -v
+ *
+ * [no remotes]
+ */
+const prepareNoRemoteGitRepository = async (): Promise<{
+  gitDir: string;
+  cleanup: () => Promise<void>;
+}> => {
+  const tempDir = await prepareTempDir();
+  await fsExec('git init test', { cwd: tempDir });
+  const gitDir = path.resolve(tempDir, 'test');
+
+  const cleanup = async () => {
+    return fsRemove(tempDir, { recursive: true });
+  };
+  return {
+    gitDir,
+    cleanup
+  };
+};
+
+/**
+ * git remote -v
+ *
+ * origin  /tmp/remote.git (fetch)
+ * origin  /tmp/remote.git (push)
+ */
+const prepareOneRemoteGitRepository = async (): Promise<{
+  gitDir: string;
+  cleanup: () => Promise<void>;
+}> => {
+  const tempDir = await prepareTempDir();
+  await fsExec('git init --bare remote.git', { cwd: tempDir });
+  await fsExec('git clone remote.git test', { cwd: tempDir });
+  const gitDir = path.resolve(tempDir, 'test');
+
+  const cleanup = async () => {
+    return fsRemove(tempDir, { recursive: true });
+  };
+  return {
+    gitDir,
+    cleanup
+  };
+};
+
+/**
+ * git remote -v
+ *
+ * origin  /tmp/remote.git (fetch)
+ * origin  /tmp/remote.git (push)
+ * other   ../remote2.git (fetch)
+ * other   ../remote2.git (push)
+ */
+const prepareTwoRemotesGitRepository = async (): Promise<{
+  gitDir: string;
+  cleanup: () => Promise<void>;
+}> => {
+  const tempDir = await prepareTempDir();
+  await fsExec('git init --bare remote.git', { cwd: tempDir });
+  await fsExec('git init --bare other.git', { cwd: tempDir });
+  await fsExec('git clone remote.git test', { cwd: tempDir });
+  const gitDir = path.resolve(tempDir, 'test');
+  await fsExec('git remote add other ../other.git', { cwd: gitDir });
+
+  const cleanup = async () => {
+    return fsRemove(tempDir, { recursive: true });
+  };
+  return {
+    gitDir,
+    cleanup
+  };
+};
+
+describe('cli flow to push git branch', () => {
+  it('do nothing when OCO_GITPUSH is set to false', async () => {
+    const { gitDir, cleanup } = await prepareNoRemoteGitRepository();
+
+    await render('echo', [`'console.log("Hello World");' > index.ts`], {
+      cwd: gitDir
+    });
+    await render('git', ['add index.ts'], { cwd: gitDir });
+
+    const { queryByText, findByText, userEvent } = await render(
+      `OCO_AI_PROVIDER='test' OCO_GITPUSH='false' node`,
+      [resolve('./out/cli.cjs')],
+      { cwd: gitDir }
+    );
+    expect(await findByText('Confirm the commit message?')).toBeInTheConsole();
+    userEvent.keyboard('[Enter]');
+
+    expect(
+      await queryByText('Choose a remote to push to')
+    ).not.toBeInTheConsole();
+    expect(
+      await queryByText('Do you want to run `git push`?')
+    ).not.toBeInTheConsole();
+    expect(
+      await queryByText('Successfully pushed all commits to origin')
+    ).not.toBeInTheConsole();
+    expect(
+      await queryByText('Command failed with exit code 1')
+    ).not.toBeInTheConsole();
+
+    await cleanup();
+  });
+
+  it('push and cause error when there is no remote', async () => {
+    const { gitDir, cleanup } = await prepareNoRemoteGitRepository();
+
+    await render('echo', [`'console.log("Hello World");' > index.ts`], {
+      cwd: gitDir
+    });
+    await render('git', ['add index.ts'], { cwd: gitDir });
+
+    const { queryByText, findByText, userEvent } = await render(
+      `OCO_AI_PROVIDER='test' node`,
+      [resolve('./out/cli.cjs')],
+      { cwd: gitDir }
+    );
+    expect(await findByText('Confirm the commit message?')).toBeInTheConsole();
+    userEvent.keyboard('[Enter]');
+
+    expect(
+      await queryByText('Choose a remote to push to')
+    ).not.toBeInTheConsole();
+    expect(
+      await queryByText('Do you want to run `git push`?')
+    ).not.toBeInTheConsole();
+    expect(
+      await queryByText('Successfully pushed all commits to origin')
+    ).not.toBeInTheConsole();
+
+    expect(
+      await findByText('Command failed with exit code 1')
+    ).toBeInTheConsole();
+
+    await cleanup();
+  });
+
+  it('push when one remote is set', async () => {
+    const { gitDir, cleanup } = await prepareOneRemoteGitRepository();
+
+    await render('echo', [`'console.log("Hello World");' > index.ts`], {
+      cwd: gitDir
+    });
+    await render('git', ['add index.ts'], { cwd: gitDir });
+
+    const { findByText, userEvent } = await render(
+      `OCO_AI_PROVIDER='test' node`,
+      [resolve('./out/cli.cjs')],
+      { cwd: gitDir }
+    );
+    expect(await findByText('Confirm the commit message?')).toBeInTheConsole();
+    userEvent.keyboard('[Enter]');
+
+    expect(
+      await findByText('Do you want to run `git push`?')
+    ).toBeInTheConsole();
+    userEvent.keyboard('[Enter]');
+
+    expect(
+      await findByText('Successfully pushed all commits to origin')
+    ).toBeInTheConsole();
+
+    await cleanup();
+  });
+
+  it('push when two remotes are set', async () => {
+    const { gitDir, cleanup } = await prepareTwoRemotesGitRepository();
+
+    await render('echo', [`'console.log("Hello World");' > index.ts`], {
+      cwd: gitDir
+    });
+    await render('git', ['add index.ts'], { cwd: gitDir });
+
+    const { findByText, userEvent } = await render(
+      `OCO_AI_PROVIDER='test' node`,
+      [resolve('./out/cli.cjs')],
+      { cwd: gitDir }
+    );
+    expect(await findByText('Confirm the commit message?')).toBeInTheConsole();
+    userEvent.keyboard('[Enter]');
+
+    expect(await findByText('Choose a remote to push to')).toBeInTheConsole();
+    userEvent.keyboard('[Enter]');
+
+    expect(
+      await findByText('Successfully pushed all commits to origin')
+    ).toBeInTheConsole();
+
+    await cleanup();
+  });
+});

test/e2e/oneFile.test.ts

@@ -17,7 +17,7 @@ it('cli flow to generate commit message for 1 new file (staged)', async () => {
   expect(await findByText('Confirm the commit message?')).toBeInTheConsole();
   userEvent.keyboard('[Enter]');
 
-  expect(await findByText('Choose a remote to push to')).toBeInTheConsole();
+  expect(await findByText('Do you want to run `git push`?')).toBeInTheConsole();
   userEvent.keyboard('[Enter]');
 
   expect(await findByText('Successfully pushed all commits to origin')).toBeInTheConsole();
@@ -46,7 +46,7 @@ it('cli flow to generate commit message for 1 changed file (not staged)', async
 
   expect(await findByText('Successfully committed')).toBeInTheConsole();
 
-  expect(await findByText('Choose a remote to push to')).toBeInTheConsole();
+  expect(await findByText('Do you want to run `git push`?')).toBeInTheConsole();
   userEvent.keyboard('[Enter]');
 
   expect(await findByText('Successfully pushed all commits to origin')).toBeInTheConsole();

test/e2e/prompt-module/commitlint.test.ts

@@ -181,9 +181,7 @@ describe('cli flow to generate commit message using @commitlint prompt-module',
       [],
       { cwd: gitDir }
     );
-    expect(
+    expect(await commitlintGet.findByText('consistency')).toBeInTheConsole();
-      await commitlintGet.findByText('[object Object]')
-    ).toBeInTheConsole();
 
     // Run 'oco' using .opencommit-commitlint
     await render('echo', [`'console.log("Hello World");' > index.ts`], {
@@ -211,7 +209,7 @@ describe('cli flow to generate commit message using @commitlint prompt-module',
     oco.userEvent.keyboard('[Enter]');
 
     expect(
-      await oco.findByText('Choose a remote to push to')
+      await oco.findByText('Do you want to run `git push`?')
     ).toBeInTheConsole();
     oco.userEvent.keyboard('[Enter]');
 

test/e2e/utils.ts

@@ -15,7 +15,7 @@ export const prepareEnvironment = async (): Promise<{
   gitDir: string;
   cleanup: () => Promise<void>;
 }> => {
-  const tempDir = await fsMakeTempDir(path.join(tmpdir(), 'opencommit-test-'));
+  const tempDir = await prepareTempDir();
   // Create a remote git repository int the temp directory. This is necessary to execute the `git push` command
   await fsExec('git init --bare remote.git', { cwd: tempDir }); 
   await fsExec('git clone remote.git test', { cwd: tempDir });
@@ -30,4 +30,8 @@ export const prepareEnvironment = async (): Promise<{
   }
 }
 
+export const prepareTempDir = async(): Promise<string> => {
+  return await fsMakeTempDir(path.join(tmpdir(), 'opencommit-test-'));
+}
+
 export const wait = (ms: number) => new Promise(resolve => setTimeout(resolve, ms));

test/unit/config.test.ts

@@ -1,8 +1,18 @@
-import { getConfig } from '../../src/commands/config';
+import { existsSync, readFileSync, rmSync } from 'fs';
+import {
+  CONFIG_KEYS,
+  DEFAULT_CONFIG,
+  getConfig,
+  setConfig
+} from '../../src/commands/config';
 import { prepareFile } from './utils';
+import { dirname } from 'path';
 
-describe('getConfig', () => {
+describe('config', () => {
   const originalEnv = { ...process.env };
+  let globalConfigFile: { filePath: string; cleanup: () => Promise<void> };
+  let envConfigFile: { filePath: string; cleanup: () => Promise<void> };
+
   function resetEnv(env: NodeJS.ProcessEnv) {
     Object.keys(process.env).forEach((key) => {
       if (!(key in env)) {
@@ -13,93 +23,281 @@ describe('getConfig', () => {
     });
   }
 
-  beforeEach(() => {
+  beforeEach(async () => {
     resetEnv(originalEnv);
+    if (globalConfigFile) await globalConfigFile.cleanup();
+    if (envConfigFile) await envConfigFile.cleanup();
+  });
+
+  afterEach(async () => {
+    if (globalConfigFile) await globalConfigFile.cleanup();
+    if (envConfigFile) await envConfigFile.cleanup();
   });
 
   afterAll(() => {
     resetEnv(originalEnv);
   });
 
-  it('return config values from the global config file', async () => {
+  const generateConfig = async (
-    const configFile = await prepareFile(
+    fileName: string,
-      '.opencommit',
+    content: Record<string, string>
-      `
+  ) => {
-OCO_OPENAI_API_KEY="sk-key"
+    const fileContent = Object.entries(content)
-OCO_ANTHROPIC_API_KEY="secret-key"
+      .map(([key, value]) => `${key}="${value}"`)
-OCO_TOKENS_MAX_INPUT="8192"
+      .join('\n');
-OCO_TOKENS_MAX_OUTPUT="1000"
+    return await prepareFile(fileName, fileContent);
-OCO_OPENAI_BASE_PATH="/openai/api"
+  };
-OCO_DESCRIPTION="true"
-OCO_EMOJI="true"
-OCO_MODEL="gpt-4"
-OCO_LANGUAGE="de"
-OCO_MESSAGE_TEMPLATE_PLACEHOLDER="$m"
-OCO_PROMPT_MODULE="@commitlint"
-OCO_AI_PROVIDER="ollama"
-OCO_GITPUSH="false"
-OCO_ONE_LINE_COMMIT="true"
-`
-    );
-    const config = getConfig({ configPath: configFile.filePath, envPath: '' });
 
-    expect(config).not.toEqual(null);
+  describe('getConfig', () => {
-    expect(config!['OCO_OPENAI_API_KEY']).toEqual('sk-key');
+    it('should prioritize local .env over global .opencommit config', async () => {
-    expect(config!['OCO_ANTHROPIC_API_KEY']).toEqual('secret-key');
+      globalConfigFile = await generateConfig('.opencommit', {
-    expect(config!['OCO_TOKENS_MAX_INPUT']).toEqual(8192);
+        OCO_API_KEY: 'global-key',
-    expect(config!['OCO_TOKENS_MAX_OUTPUT']).toEqual(1000);
+        OCO_MODEL: 'gpt-3.5-turbo',
-    expect(config!['OCO_OPENAI_BASE_PATH']).toEqual('/openai/api');
+        OCO_LANGUAGE: 'en'
-    expect(config!['OCO_DESCRIPTION']).toEqual(true);
+      });
-    expect(config!['OCO_EMOJI']).toEqual(true);
-    expect(config!['OCO_MODEL']).toEqual('gpt-4');
-    expect(config!['OCO_LANGUAGE']).toEqual('de');
-    expect(config!['OCO_MESSAGE_TEMPLATE_PLACEHOLDER']).toEqual('$m');
-    expect(config!['OCO_PROMPT_MODULE']).toEqual('@commitlint');
-    expect(() => ['ollama', 'gemini'].includes(config!['OCO_AI_PROVIDER'])).toBeTruthy();
-    expect(config!['OCO_GITPUSH']).toEqual(false);
-    expect(config!['OCO_ONE_LINE_COMMIT']).toEqual(true);
 
-    await configFile.cleanup();
+      envConfigFile = await generateConfig('.env', {
+        OCO_API_KEY: 'local-key',
+        OCO_LANGUAGE: 'fr'
+      });
+
+      const config = getConfig({
+        globalPath: globalConfigFile.filePath,
+        envPath: envConfigFile.filePath
+      });
+
+      expect(config).not.toEqual(null);
+      expect(config.OCO_API_KEY).toEqual('local-key');
+      expect(config.OCO_MODEL).toEqual('gpt-3.5-turbo');
+      expect(config.OCO_LANGUAGE).toEqual('fr');
+    });
+
+    it('should fallback to global config when local config is not set', async () => {
+      globalConfigFile = await generateConfig('.opencommit', {
+        OCO_API_KEY: 'global-key',
+        OCO_MODEL: 'gpt-4',
+        OCO_LANGUAGE: 'de',
+        OCO_DESCRIPTION: 'true'
+      });
+
+      envConfigFile = await generateConfig('.env', {
+        OCO_API_URL: 'local-api-url'
+      });
+
+      const config = getConfig({
+        globalPath: globalConfigFile.filePath,
+        envPath: envConfigFile.filePath
+      });
+
+      expect(config).not.toEqual(null);
+      expect(config.OCO_API_KEY).toEqual('global-key');
+      expect(config.OCO_API_URL).toEqual('local-api-url');
+      expect(config.OCO_MODEL).toEqual('gpt-4');
+      expect(config.OCO_LANGUAGE).toEqual('de');
+      expect(config.OCO_DESCRIPTION).toEqual(true);
+    });
+
+    it('should handle boolean and numeric values correctly', async () => {
+      globalConfigFile = await generateConfig('.opencommit', {
+        OCO_TOKENS_MAX_INPUT: '4096',
+        OCO_TOKENS_MAX_OUTPUT: '500',
+        OCO_GITPUSH: 'true'
+      });
+
+      envConfigFile = await generateConfig('.env', {
+        OCO_TOKENS_MAX_INPUT: '8192',
+        OCO_ONE_LINE_COMMIT: 'false'
+      });
+
+      const config = getConfig({
+        globalPath: globalConfigFile.filePath,
+        envPath: envConfigFile.filePath
+      });
+
+      expect(config).not.toEqual(null);
+      expect(config.OCO_TOKENS_MAX_INPUT).toEqual(8192);
+      expect(config.OCO_TOKENS_MAX_OUTPUT).toEqual(500);
+      expect(config.OCO_GITPUSH).toEqual(true);
+      expect(config.OCO_ONE_LINE_COMMIT).toEqual(false);
+    });
+
+    it('should handle empty local config correctly', async () => {
+      globalConfigFile = await generateConfig('.opencommit', {
+        OCO_API_KEY: 'global-key',
+        OCO_MODEL: 'gpt-4',
+        OCO_LANGUAGE: 'es'
+      });
+
+      envConfigFile = await generateConfig('.env', {});
+
+      const config = getConfig({
+        globalPath: globalConfigFile.filePath,
+        envPath: envConfigFile.filePath
+      });
+
+      expect(config).not.toEqual(null);
+      expect(config.OCO_API_KEY).toEqual('global-key');
+      expect(config.OCO_MODEL).toEqual('gpt-4');
+      expect(config.OCO_LANGUAGE).toEqual('es');
+    });
+
+    it('should override global config with null values in local .env', async () => {
+      globalConfigFile = await generateConfig('.opencommit', {
+        OCO_API_KEY: 'global-key',
+        OCO_MODEL: 'gpt-4',
+        OCO_LANGUAGE: 'es'
+      });
+
+      envConfigFile = await generateConfig('.env', {
+        OCO_API_KEY: 'null'
+      });
+
+      const config = getConfig({
+        globalPath: globalConfigFile.filePath,
+        envPath: envConfigFile.filePath
+      });
+
+      expect(config).not.toEqual(null);
+      expect(config.OCO_API_KEY).toEqual(null);
+    });
+
+    it('should handle empty global config', async () => {
+      globalConfigFile = await generateConfig('.opencommit', {});
+      envConfigFile = await generateConfig('.env', {});
+
+      const config = getConfig({
+        globalPath: globalConfigFile.filePath,
+        envPath: envConfigFile.filePath
+      });
+
+      expect(config).not.toEqual(null);
+      expect(config.OCO_API_KEY).toEqual(undefined);
+    });
   });
 
-  it('return config values from the local env file', async () => {
+  describe('setConfig', () => {
-    const envFile = await prepareFile(
+    beforeEach(async () => {
-      '.env',
+      // we create and delete the file to have the parent directory, but not the file, to test the creation of the file
-      `
+      globalConfigFile = await generateConfig('.opencommit', {});
-OCO_OPENAI_API_KEY="sk-key"
+      rmSync(globalConfigFile.filePath);
-OCO_ANTHROPIC_API_KEY="secret-key"
+    });
-OCO_TOKENS_MAX_INPUT="8192"
-OCO_TOKENS_MAX_OUTPUT="1000"
-OCO_OPENAI_BASE_PATH="/openai/api"
-OCO_DESCRIPTION="true"
-OCO_EMOJI="true"
-OCO_MODEL="gpt-4"
-OCO_LANGUAGE="de"
-OCO_MESSAGE_TEMPLATE_PLACEHOLDER="$m"
-OCO_PROMPT_MODULE="@commitlint"
-OCO_AI_PROVIDER="ollama"
-OCO_GITPUSH="false"
-OCO_ONE_LINE_COMMIT="true"
-    `
-    );
-    const config = getConfig({ configPath: '', envPath: envFile.filePath });
 
-    expect(config).not.toEqual(null);
+    it('should create .opencommit file with DEFAULT CONFIG if it does not exist on first setConfig run', async () => {
-    expect(config!['OCO_OPENAI_API_KEY']).toEqual('sk-key');
+      const isGlobalConfigFileExist = existsSync(globalConfigFile.filePath);
-    expect(config!['OCO_ANTHROPIC_API_KEY']).toEqual('secret-key');
+      expect(isGlobalConfigFileExist).toBe(false);
-    expect(config!['OCO_TOKENS_MAX_INPUT']).toEqual(8192);
-    expect(config!['OCO_TOKENS_MAX_OUTPUT']).toEqual(1000);
-    expect(config!['OCO_OPENAI_BASE_PATH']).toEqual('/openai/api');
-    expect(config!['OCO_DESCRIPTION']).toEqual(true);
-    expect(config!['OCO_EMOJI']).toEqual(true);
-    expect(config!['OCO_MODEL']).toEqual('gpt-4');
-    expect(config!['OCO_LANGUAGE']).toEqual('de');
-    expect(config!['OCO_MESSAGE_TEMPLATE_PLACEHOLDER']).toEqual('$m');
-    expect(config!['OCO_PROMPT_MODULE']).toEqual('@commitlint');
-    expect(() => ['ollama', 'gemini'].includes(config!['OCO_AI_PROVIDER'])).toBeTruthy();
-    expect(config!['OCO_GITPUSH']).toEqual(false);
-    expect(config!['OCO_ONE_LINE_COMMIT']).toEqual(true);
 
-    await envFile.cleanup();
+      await setConfig(
+        [[CONFIG_KEYS.OCO_API_KEY, 'persisted-key_1']],
+        globalConfigFile.filePath
+      );
+
+      const fileContent = readFileSync(globalConfigFile.filePath, 'utf8');
+      expect(fileContent).toContain('OCO_API_KEY=persisted-key_1');
+      Object.entries(DEFAULT_CONFIG).forEach(([key, value]) => {
+        expect(fileContent).toContain(`${key}=${value}`);
+      });
+    });
+
+    it('should set new config values', async () => {
+      globalConfigFile = await generateConfig('.opencommit', {});
+      await setConfig(
+        [
+          [CONFIG_KEYS.OCO_API_KEY, 'new-key'],
+          [CONFIG_KEYS.OCO_MODEL, 'gpt-4']
+        ],
+        globalConfigFile.filePath
+      );
+
+      const config = getConfig({
+        globalPath: globalConfigFile.filePath
+      });
+      expect(config.OCO_API_KEY).toEqual('new-key');
+      expect(config.OCO_MODEL).toEqual('gpt-4');
+    });
+
+    it('should update existing config values', async () => {
+      globalConfigFile = await generateConfig('.opencommit', {
+        OCO_API_KEY: 'initial-key'
+      });
+      await setConfig(
+        [[CONFIG_KEYS.OCO_API_KEY, 'updated-key']],
+        globalConfigFile.filePath
+      );
+
+      const config = getConfig({
+        globalPath: globalConfigFile.filePath
+      });
+      expect(config.OCO_API_KEY).toEqual('updated-key');
+    });
+
+    it('should handle boolean and numeric values correctly', async () => {
+      globalConfigFile = await generateConfig('.opencommit', {});
+      await setConfig(
+        [
+          [CONFIG_KEYS.OCO_TOKENS_MAX_INPUT, '8192'],
+          [CONFIG_KEYS.OCO_DESCRIPTION, 'true'],
+          [CONFIG_KEYS.OCO_ONE_LINE_COMMIT, 'false']
+        ],
+        globalConfigFile.filePath
+      );
+
+      const config = getConfig({
+        globalPath: globalConfigFile.filePath
+      });
+      expect(config.OCO_TOKENS_MAX_INPUT).toEqual(8192);
+      expect(config.OCO_DESCRIPTION).toEqual(true);
+      expect(config.OCO_ONE_LINE_COMMIT).toEqual(false);
+    });
+
+    it('should throw an error for unsupported config keys', async () => {
+      globalConfigFile = await generateConfig('.opencommit', {});
+
+      try {
+        await setConfig(
+          [['UNSUPPORTED_KEY', 'value']],
+          globalConfigFile.filePath
+        );
+        throw new Error('NEVER_REACHED');
+      } catch (error) {
+        expect(error.message).toContain(
+          'Unsupported config key: UNSUPPORTED_KEY'
+        );
+        expect(error.message).not.toContain('NEVER_REACHED');
+      }
+    });
+
+    it('should persist changes to the config file', async () => {
+      const isGlobalConfigFileExist = existsSync(globalConfigFile.filePath);
+      expect(isGlobalConfigFileExist).toBe(false);
+
+      await setConfig(
+        [[CONFIG_KEYS.OCO_API_KEY, 'persisted-key']],
+        globalConfigFile.filePath
+      );
+
+      const fileContent = readFileSync(globalConfigFile.filePath, 'utf8');
+      expect(fileContent).toContain('OCO_API_KEY=persisted-key');
+    });
+
+    it('should set multiple configs in a row and keep the changes', async () => {
+      const isGlobalConfigFileExist = existsSync(globalConfigFile.filePath);
+      expect(isGlobalConfigFileExist).toBe(false);
+
+      await setConfig(
+        [[CONFIG_KEYS.OCO_API_KEY, 'persisted-key']],
+        globalConfigFile.filePath
+      );
+
+      const fileContent1 = readFileSync(globalConfigFile.filePath, 'utf8');
+      expect(fileContent1).toContain('OCO_API_KEY=persisted-key');
+
+      await setConfig(
+        [[CONFIG_KEYS.OCO_MODEL, 'gpt-4']],
+        globalConfigFile.filePath
+      );
+
+      const fileContent2 = readFileSync(globalConfigFile.filePath, 'utf8');
+      expect(fileContent2).toContain('OCO_MODEL=gpt-4');
+    });
   });
-});
+});

test/unit/gemini.test.ts

@@ -1,80 +1,69 @@
-import { Gemini } from '../../src/engine/gemini';
+import { GeminiEngine } from '../../src/engine/gemini';
-import { ChatCompletionRequestMessage } from 'openai';
+
 import { GenerativeModel, GoogleGenerativeAI } from '@google/generative-ai';
-import { ConfigType, getConfig } from '../../src/commands/config';
+import {
+  ConfigType,
+  getConfig,
+  OCO_AI_PROVIDER_ENUM
+} from '../../src/commands/config';
+import { OpenAI } from 'openai';
 
 describe('Gemini', () => {
-  let gemini: Gemini;
+  let gemini: GeminiEngine;
   let mockConfig: ConfigType;
   let mockGoogleGenerativeAi: GoogleGenerativeAI;
   let mockGenerativeModel: GenerativeModel;
   let mockExit: jest.SpyInstance<never, [code?: number | undefined], any>;
-  let mockWarmup: jest.SpyInstance<any, unknown[], any>;
+
-  
+  const noop: (...args: any[]) => any = (...args: any[]) => {};
-  const noop: (code?: number | undefined) => never = (code?: number | undefined) => {};
+
-  
   const mockGemini = () => {
-    gemini = new Gemini();
+    mockConfig = getConfig() as ConfigType;
-  }
+
-  
+    gemini = new GeminiEngine({
+      apiKey: mockConfig.OCO_API_KEY,
+      model: mockConfig.OCO_MODEL
+    });
+  };
+
   const oldEnv = process.env;
 
   beforeEach(() => {
     jest.resetModules();
     process.env = { ...oldEnv };
-    
+
     jest.mock('@google/generative-ai');
     jest.mock('../src/commands/config');
-    
+
     jest.mock('@clack/prompts', () => ({
       intro: jest.fn(),
-      outro: jest.fn(),
+      outro: jest.fn()
     }));
-    
+
-    if (mockWarmup) mockWarmup.mockRestore();
-    
     mockExit = jest.spyOn(process, 'exit').mockImplementation();
+
     mockConfig = getConfig() as ConfigType;
-    
+
-    mockConfig.OCO_AI_PROVIDER = 'gemini';
+    mockConfig.OCO_AI_PROVIDER = OCO_AI_PROVIDER_ENUM.GEMINI;
-    mockConfig.OCO_GEMINI_API_KEY = 'mock-api-key';
+    mockConfig.OCO_API_KEY = 'mock-api-key';
     mockConfig.OCO_MODEL = 'gemini-1.5-flash';
-    
+
-    mockGoogleGenerativeAi = new GoogleGenerativeAI(mockConfig.OCO_GEMINI_API_KEY);    
+    mockGoogleGenerativeAi = new GoogleGenerativeAI(mockConfig.OCO_API_KEY);
-    mockGenerativeModel = mockGoogleGenerativeAi.getGenerativeModel({ model: mockConfig.OCO_MODEL, });
+    mockGenerativeModel = mockGoogleGenerativeAi.getGenerativeModel({
+      model: mockConfig.OCO_MODEL
+    });
   });
-  
+
   afterEach(() => {
     gemini = undefined as any;
-  })
+  });
-  
+
   afterAll(() => {
     mockExit.mockRestore();
     process.env = oldEnv;
   });
 
-  it('should initialize with correct config', () => {
+  it.skip('should exit process if OCO_GEMINI_API_KEY is not set and command is not config', () => {
-    mockGemini();
-    // gemini = new Gemini();
-    expect(gemini).toBeDefined();
-  });
-
-  it('should warmup correctly', () => {
-    mockWarmup = jest.spyOn(Gemini.prototype as any, 'warmup').mockImplementation(noop);
-    mockGemini();
-    expect(gemini).toBeDefined();
-  });
-
-  it('should exit process if OCO_GEMINI_API_KEY is not set and command is not config', () => {
-    process.env.OCO_GEMINI_API_KEY = undefined;
-    process.env.OCO_AI_PROVIDER = 'gemini';
-    
-    mockGemini();
-    
-    expect(mockExit).toHaveBeenCalledWith(1);
-  });
-
-  it('should exit process if model is not supported and command is not config', () => {
     process.env.OCO_GEMINI_API_KEY = undefined;
     process.env.OCO_AI_PROVIDER = 'gemini';
 
@@ -82,24 +71,26 @@ describe('Gemini', () => {
 
     expect(mockExit).toHaveBeenCalledWith(1);
   });
-  
+
   it('should generate commit message', async () => {
-    const mockGenerateContent = jest.fn().mockResolvedValue({ response: { text: () => 'generated content' } });
+    const mockGenerateContent = jest
+      .fn()
+      .mockResolvedValue({ response: { text: () => 'generated content' } });
     mockGenerativeModel.generateContent = mockGenerateContent;
-    
+
-    mockWarmup = jest.spyOn(Gemini.prototype as any, 'warmup').mockImplementation(noop);
     mockGemini();
-    
+
-    const messages: ChatCompletionRequestMessage[] = [
+    const messages: Array<OpenAI.Chat.Completions.ChatCompletionMessageParam> =
-      { role: 'system', content: 'system message' },
+      [
-      { role: 'assistant', content: 'assistant message' },
+        { role: 'system', content: 'system message' },
-    ];
+        { role: 'assistant', content: 'assistant message' }
-    
+      ];
-    jest.spyOn(gemini, 'generateCommitMessage').mockImplementation(async () => 'generated content');
+
+    jest
+      .spyOn(gemini, 'generateCommitMessage')
+      .mockImplementation(async () => 'generated content');
     const result = await gemini.generateCommitMessage(messages);
 
     expect(result).toEqual('generated content');
-    expect(mockWarmup).toHaveBeenCalled();
   });
-  
+});
-});

test/unit/utils.ts

@@ -1,7 +1,7 @@
-import path from 'path';
+import { existsSync, mkdtemp, rm, writeFile } from 'fs';
-import { mkdtemp, rm, writeFile } from 'fs';
-import { promisify } from 'util';
 import { tmpdir } from 'os';
+import path from 'path';
+import { promisify } from 'util';
 const fsMakeTempDir = promisify(mkdtemp);
 const fsRemove = promisify(rm);
 const fsWriteFile = promisify(writeFile);
@@ -20,8 +20,11 @@ export async function prepareFile(
   const filePath = path.resolve(tempDir, fileName);
   await fsWriteFile(filePath, content);
   const cleanup = async () => {
-    return fsRemove(tempDir, { recursive: true });
+    if (existsSync(tempDir)) {
+      await fsRemove(tempDir, { recursive: true });
+    }
   };
+
   return {
     filePath,
     cleanup

tsconfig.json

@@ -4,7 +4,7 @@
     "lib": ["ES6", "ES2020"],
 
     "module": "CommonJS",
-    
+
     "resolveJsonModule": true,
     "moduleResolution": "Node",
 
@@ -21,9 +21,7 @@
 
     "skipLibCheck": true
   },
-  "include": [
+  "include": ["test/jest-setup.ts"],
-    "test/jest-setup.ts"
-  ],
   "exclude": ["node_modules"],
   "ts-node": {
     "esm": true,