@negoziator/ai-commit 2.9.2 → 2.13.0

package/README.md

@@ -15,8 +15,15 @@
 - [Usage](#usage)
     - [CLI Mode](#cli-mode)
     - [Git Hook Integration](#git-hook-integration)
+- [LLM Providers](#llm-providers)
+    - [OpenAI (Default)](#openai-default)
+    - [Anthropic Claude](#anthropic-claude)
+    - [Azure OpenAI](#azure-openai)
+    - [Ollama (Local)](#ollama-local)
+    - [Custom/RAG](#customrag)
 - [Configuration](#configuration)
     - [Options](#options)
+    - [Environment Variables (.env)](#environment-variables-env)
     - [Project-Specific Configuration](#project-specific-configuration)
 - [Maintainers](#maintainers)
 - [Contributing](#contributing)
@@ -29,6 +36,7 @@
 - 🌐 **Multiple Languages**: Supports commit messages in different locales
 - ⚙️ **Customizable**: Configure the AI model, message length, and other parameters
 - 📝 **Project Context**: Add project-specific context to improve message relevance
+- 🔌 **Multiple LLM Providers**: Choose from OpenAI, Anthropic, Azure OpenAI, Ollama, or custom endpoints
 
 ## Setup
 > A minimum of Node v20 is required. Check your Node.js version with `node --version`.
@@ -39,17 +47,37 @@
    npm install -g @negoziator/ai-commit
    ```
 
-2. **Retrieve your API key from [OpenAI](https://platform.openai.com/account/api-keys)**
+2. **Choose and configure your LLM provider:**
 
-   > Note: This requires an OpenAI account. If you don't have one, you can sign up for a free trial.
+   **Option A: OpenAI (Default)**
+   ```sh
+   aicommit config set OPENAI_KEY=<your-api-key>
+   ```
+   Get your API key from [OpenAI Platform](https://platform.openai.com/account/api-keys)
+
+   **Option B: Anthropic Claude**
+   ```sh
+   aicommit config set provider=anthropic
+   aicommit config set ANTHROPIC_KEY=<your-api-key>
+   ```
+   Get your API key from [Anthropic Console](https://console.anthropic.com/)
 
-3. **Set the key so aicommit can use it:**
+   **Option C: Ollama (Local, Free)**
+   ```sh
+   # Install Ollama from https://ollama.com, then:
+   ollama pull llama3.2
+   aicommit config set provider=ollama
+   aicommit config set model=llama3.2
+   ```
 
+   **Option D: Azure OpenAI**
    ```sh
-   aicommit config set OPENAI_KEY=<your token>
+   aicommit config set provider=azure-openai
+   aicommit config set AZURE_OPENAI_KEY=<your-key>
+   aicommit config set AZURE_ENDPOINT=<your-endpoint>
    ```
 
-   This will create a `.aicommit` file in your home directory.
+   See the [LLM Providers](#llm-providers) section for more options and details.
 
 ### Upgrading
 
@@ -98,6 +126,125 @@ To uninstall the hook:
 aicommit hook uninstall
 ```
 
+## LLM Providers
+
+AI-Commit supports multiple LLM providers, allowing you to choose the best option for your needs.
+
+### OpenAI (Default)
+
+The default provider using OpenAI's GPT models.
+
+**Setup:**
+```sh
+aicommit config set OPENAI_KEY=<your-api-key>
+```
+
+**Recommended models:**
+- `gpt-4o-mini` (default, fast and cost-effective)
+- `gpt-4o` (more capable, higher cost)
+- `gpt-4-turbo`
+
+**Get your API key:** [OpenAI Platform](https://platform.openai.com/account/api-keys)
+
+---
+
+### Anthropic Claude
+
+Use Anthropic's Claude models as an alternative to OpenAI.
+
+**Setup:**
+```sh
+aicommit config set provider=anthropic
+aicommit config set ANTHROPIC_KEY=<your-api-key>
+```
+
+**Recommended models:**
+- `claude-3-5-sonnet-20241022` (recommended, best balance)
+- `claude-3-opus-20240229` (most capable)
+- `claude-3-haiku-20240307` (fastest, most economical)
+
+**Get your API key:** [Anthropic Console](https://console.anthropic.com/)
+
+---
+
+### Azure OpenAI
+
+Use Azure's OpenAI Service for enterprise deployments.
+
+**Setup:**
+```sh
+aicommit config set provider=azure-openai
+aicommit config set AZURE_OPENAI_KEY=<your-api-key>
+aicommit config set AZURE_ENDPOINT=<your-endpoint>
+```
+
+**Example endpoint:** `https://your-resource.openai.azure.com`
+
+**Note:** The `model` config should match your Azure deployment name.
+
+**Learn more:** [Azure OpenAI Service](https://azure.microsoft.com/en-us/products/ai-services/openai-service)
+
+---
+
+### Ollama (Local)
+
+Run AI-Commit completely offline using local models via Ollama.
+
+**Setup:**
+```sh
+# 1. Install Ollama from https://ollama.com
+# 2. Pull a model
+ollama pull llama3.2
+
+# 3. Configure AI-Commit
+aicommit config set provider=ollama
+aicommit config set model=llama3.2
+```
+
+**Recommended models:**
+- `llama3.2` (recommended, good balance)
+- `codellama` (optimized for code)
+- `mistral` (fast and capable)
+- `qwen2.5-coder` (specialized for coding)
+
+**Default endpoint:** `http://localhost:11434` (automatically configured)
+
+**Benefits:**
+- ✅ Completely free
+- ✅ Works offline
+- ✅ Privacy-focused (data never leaves your machine)
+- ✅ No API key required
+
+---
+
+### Custom/RAG
+
+Connect to custom LLM endpoints, RAG systems, or OpenAI-compatible APIs.
+
+**Setup:**
+```sh
+aicommit config set provider=custom
+aicommit config set CUSTOM_ENDPOINT=<your-endpoint-url>
+aicommit config set CUSTOM_KEY=<optional-api-key>
+```
+
+**Compatible with:**
+- Custom RAG implementations
+- LM Studio
+- LocalAI
+- Text Generation WebUI
+- vLLM
+- Any OpenAI-compatible API
+
+**Example:**
+```sh
+aicommit config set provider=custom
+aicommit config set CUSTOM_ENDPOINT=https://my-rag.example.com/v1/chat/completions
+aicommit config set model=my-custom-model
+```
+
+---
+
 ## Configuration
 Manage configuration using the `aicommit config` command.
 
@@ -121,33 +268,103 @@ aicommit config set <key>=<value>
 
 ### Options
 
+#### General Options
+
 | Option              | Default         | Description                                                                           |
 |---------------------|-----------------|---------------------------------------------------------------------------------------|
-| `OPENAI_KEY`        | N/A             | The OpenAI API key.                                                                   |
-| `locale`            | `en`            | Locale for the generated commit messages.                                             |
-| `generate`          | `1`             | Number of commit messages to generate.                                                |
-| `model`             | `gpt-4o-mini`   | The Chat Completions model to use.                                                    |
-| `timeout`           | `10000` ms      | Network request timeout to the OpenAI API.                                            |
-| `max-length`        | `50`            | Maximum character length of the generated commit message.                             |
-| `type`              | `""`            | Type of commit message to generate.                                                   |
-| `auto-confirm`      | `false`         | Automatically confirm the generated commit message without user prompt.               |
-| `prepend-reference` | `false`         | Prepend issue reference from branch name to commit message.                           |
-| `temperature`       | `0.2`           | The temperature (0.0-2.0) is used to control the randomness of the output from OpenAI |
-| `max-completion-tokens` | `10000`         | Maximum number of tokens that can be generated in the completion                      |
+| `provider`          | `openai`        | LLM provider to use (`openai`, `anthropic`, `azure-openai`, `ollama`, `custom`)      |
+| `locale`            | `en`            | Locale for the generated commit messages                                              |
+| `generate`          | `1`             | Number of commit messages to generate                                                 |
+| `model`             | `gpt-4o-mini`   | The model to use (provider-specific)                                                  |
+| `timeout`           | `10000`         | Network request timeout in milliseconds                                               |
+| `max-length`        | `50`            | Maximum character length of the generated commit message                              |
+| `type`              | `""`            | Type of commit message to generate (`conventional` or empty)                          |
+| `auto-confirm`      | `false`         | Automatically confirm the generated commit message without user prompt                |
+| `prepend-reference` | `false`         | Prepend issue reference from branch name to commit message                            |
+| `temperature`       | `0.2`           | Temperature (0.0-2.0) to control randomness of the output                             |
+| `max-completion-tokens` | `10000`     | Maximum number of tokens that can be generated in the completion                      |
+| `proxy`             | N/A             | HTTPS proxy URL (e.g., `http://proxy.example.com:8080`)                              |
+
+#### Provider-Specific Options
+
+| Option              | Provider        | Description                                                                           |
+|---------------------|-----------------|---------------------------------------------------------------------------------------|
+| `OPENAI_KEY`        | OpenAI          | OpenAI API key (starts with `sk-`)                                                    |
+| `ANTHROPIC_KEY`     | Anthropic       | Anthropic API key (starts with `sk-ant-`)                                             |
+| `AZURE_OPENAI_KEY`  | Azure OpenAI    | Azure OpenAI API key                                                                  |
+| `AZURE_ENDPOINT`    | Azure OpenAI    | Azure OpenAI endpoint URL (e.g., `https://your-resource.openai.azure.com`)           |
+| `OLLAMA_ENDPOINT`   | Ollama          | Ollama server endpoint (default: `http://localhost:11434`)                            |
+| `CUSTOM_ENDPOINT`   | Custom          | Custom API endpoint URL                                                               |
+| `CUSTOM_KEY`        | Custom          | Custom API key (optional, for endpoints requiring authentication)                     |
+
+### Environment Variables (.env)
+
+For local development, you can create environment files in your project root:
+
+**For CLI usage** (`.env`):
+```bash
+# .env - Used when running aicommit
+provider=anthropic
+ANTHROPIC_KEY=sk-ant-...
+model=claude-3-5-sonnet-20241022
+max-length=80
+```
+
+**For testing** (`.env.local`):
+```bash
+# .env.local - Used when running npm test
+OPENAI_KEY=sk-...
+# or
+provider=anthropic
+ANTHROPIC_KEY=sk-ant-...
+```
+
+**File usage:**
+- `.env` → Loaded when running `aicommit` (CLI usage)
+- `.env.local` → Loaded when running `npm test` (local testing only)
+- Both files are gitignored for security
+
+**Priority order** (highest to lowest):
+1. `.ai-commit.json` (project-specific)
+2. CLI arguments
+3. Environment variables (`.env`, `.env.local`, or shell)
+4. Global config (`~/.aicommit`)
+
+**Note:** See `.env.example` for all available options. Copy it to `.env` for CLI usage or `.env.local` for testing.
 
 ### Project-Specific Configuration
 
 You can add a `.ai-commit.json` file in the root of your project to provide additional context about your project to the AI and to override global configuration settings for the specific project.
 
-Example `.ai-commit.json`:
+**Example with OpenAI:**
 ```json
 {
-  "projectPrompt": "This is a Node.js CLI tool that uses OpenAI to generate meaningful git commit messages.",
-  "model": "gpt-4",
+  "projectPrompt": "This is a Node.js CLI tool that uses AI to generate meaningful git commit messages.",
+  "model": "gpt-4o",
   "locale": "en",
   "max-length": "100",
-  "temperature": "0.5",
-  "max-completion-tokens": "5000"
+  "temperature": "0.5"
+}
+```
+
+**Example with Anthropic:**
+```json
+{
+  "provider": "anthropic",
+  "ANTHROPIC_KEY": "sk-ant-...",
+  "model": "claude-3-5-sonnet-20241022",
+  "projectPrompt": "This is a TypeScript library for data validation.",
+  "max-length": "80"
+}
+```
+
+**Example with Ollama:**
+```json
+{
+  "provider": "ollama",
+  "model": "codellama",
+  "projectPrompt": "This is a Python web application using FastAPI.",
+  "temperature": "0.3"
 }
 ```
 

package/dist/anthropic-DJVOPReW.mjs

@@ -0,0 +1,5 @@
+var y=Object.defineProperty;var u=(f,t)=>y(f,"name",{value:t,configurable:!0});import l from"https";import{c as w}from"./index-CQ8E6Ymc.mjs";import{L as A,g as C}from"./prompt-BMR92Sng.mjs";import{K as a}from"./cli-D2zun-19.mjs";import"net";import"tls";import"url";import"assert";import"tty";import"util";import"os";import"events";import"node:buffer";import"node:path";import"node:child_process";import"node:process";import"child_process";import"path";import"fs";import"node:url";import"node:os";import"node:fs";import"buffer";import"stream";import"node:util";import"node:readline";import"node:stream";import"fs/promises";const h=class h extends A{get name(){return"anthropic"}validateConfig(){const{apiKey:t}=this.config;if(!t)throw new a("Please set your Anthropic API key via `aicommit config set ANTHROPIC_KEY=<your token>`\nGet your API key from: https://console.anthropic.com/");if(!t.startsWith("sk-ant-"))throw new a('Invalid Anthropic API key: Must start with "sk-ant-"')}async generateCommitMessage(t){try{const i=C(this.config.locale,this.config.maxLength,this.config.type,t.projectConfig),o=[{role:"user",content:t.diff}],e=[];for(let s=0;s<t.completions;s++){const r=(await this.createMessage(i,o)).content[0]?.text;r&&e.push(this.sanitizeMessage(r))}return{messages:this.deduplicateMessages(e)}}catch(i){const o=i;throw o.code==="ENOTFOUND"?new a(`Error connecting to ${o.hostname} (${o.syscall}). Are you connected to the internet?`):o}}async createMessage(t,i){const o={model:this.config.model,max_tokens:this.config.maxCompletionTokens,temperature:this.config.temperature,system:t,messages:i},{response:e,data:n}=await this.httpsPost("api.anthropic.com","/v1/messages",{"x-api-key":this.config.apiKey,"anthropic-version":"2023-06-01"},o);if(!e.statusCode||e.statusCode<200||e.statusCode>299){let s=`Anthropic API Error: ${e.statusCode} - ${e.statusMessage}`;throw n&&(s+=`
+
+${n}`),e.statusCode===500&&(s+=`
+
+Check the API status: https://status.anthropic.com`),new a(s)}return JSON.parse(n)}async httpsPost(t,i,o,e){return new Promise((n,s)=>{const c=JSON.stringify(e),r=l.request({hostname:t,path:i,method:"POST",headers:{...o,"Content-Type":"application/json","Content-Length":Buffer.byteLength(c)},timeout:this.config.timeout,agent:this.config.proxy?w(this.config.proxy):void 0},m=>{const g=[];m.on("data",d=>g.push(d)),m.on("end",()=>{n({request:r,response:m,data:Buffer.concat(g).toString()})})});r.on("error",s),r.on("timeout",()=>{r.destroy(),s(new a(`Time out error: request took over ${this.config.timeout}ms. Try increasing the \`timeout\` config, or checking the Anthropic API status https://status.anthropic.com`))}),r.write(c),r.end()})}sanitizeMessage(t){return t.trim().replace(/[\n\r]/g,"").replace(/(\w)\.$/,"$1")}deduplicateMessages(t){return Array.from(new Set(t))}};u(h,"AnthropicProvider");let p=h;export{p as default};

package/dist/azure-openai-tXzyzZc3.mjs

@@ -0,0 +1,3 @@
+var d=Object.defineProperty;var f=(u,t)=>d(u,"name",{value:t,configurable:!0});import y from"https";import{c as w}from"./index-CQ8E6Ymc.mjs";import{L as C,g as A}from"./prompt-BMR92Sng.mjs";import{K as a}from"./cli-D2zun-19.mjs";import"net";import"tls";import"url";import"assert";import"tty";import"util";import"os";import"events";import"node:buffer";import"node:path";import"node:child_process";import"node:process";import"child_process";import"path";import"fs";import"node:url";import"node:os";import"node:fs";import"buffer";import"stream";import"node:util";import"node:readline";import"node:stream";import"fs/promises";const g=class g extends C{get name(){return"azure-openai"}validateConfig(){const{apiKey:t,endpoint:n}=this.config;if(!t)throw new a("Please set your Azure OpenAI API key via `aicommit config set AZURE_OPENAI_KEY=<your token>`");if(!n)throw new a("Please set your Azure OpenAI endpoint via `aicommit config set AZURE_ENDPOINT=<your endpoint>`")}async generateCommitMessage(t){try{const n=await this.createChatCompletion({model:this.config.model,messages:[{role:"system",content:A(this.config.locale,this.config.maxLength,this.config.type,t.projectConfig)},{role:"user",content:t.diff}],temperature:this.config.temperature,top_p:1,frequency_penalty:0,presence_penalty:0,max_tokens:this.config.maxCompletionTokens,n:t.completions});return{messages:this.deduplicateMessages(n.choices.filter(r=>r.message?.content).map(r=>this.sanitizeMessage(r.message.content)))}}catch(n){const e=n;throw e.code==="ENOTFOUND"?new a(`Error connecting to ${e.hostname} (${e.syscall}). Are you connected to the internet?`):e}}async createChatCompletion(t){const e=new URL(this.config.endpoint).hostname,m=`/openai/deployments/${this.config.deploymentName||this.config.model}/chat/completions?api-version=2024-02-01`,{response:i,data:s}=await this.httpsPost(e,m,{"api-key":this.config.apiKey},t);if(!i.statusCode||i.statusCode<200||i.statusCode>299){let o=`Azure OpenAI API Error: ${i.statusCode} - ${i.statusMessage}`;throw s&&(o+=`
+
+${s}`),new a(o)}return JSON.parse(s)}async httpsPost(t,n,e,r){return new Promise((m,i)=>{const s=JSON.stringify(r),o=y.request({hostname:t,path:n,method:"POST",headers:{...e,"Content-Type":"application/json","Content-Length":Buffer.byteLength(s)},timeout:this.config.timeout,agent:this.config.proxy?w(this.config.proxy):void 0},p=>{const h=[];p.on("data",l=>h.push(l)),p.on("end",()=>{m({request:o,response:p,data:Buffer.concat(h).toString()})})});o.on("error",i),o.on("timeout",()=>{o.destroy(),i(new a(`Time out error: request took over ${this.config.timeout}ms. Try increasing the \`timeout\` config`))}),o.write(s),o.end()})}sanitizeMessage(t){return t.trim().replace(/[\n\r]/g,"").replace(/(\w)\.$/,"$1")}deduplicateMessages(t){return Array.from(new Set(t))}};f(g,"AzureOpenAIProvider");let c=g;export{c as default};