[Feature]: Adding configurable llmservice endpoint which assumes backend stores the system prompt

src/commands/config.ts

@@ -34,6 +34,8 @@ export enum CONFIG_KEYS {
   OCO_TEST_MOCK_TYPE = 'OCO_TEST_MOCK_TYPE',
   OCO_API_URL = 'OCO_API_URL',
   OCO_OLLAMA_API_URL = 'OCO_OLLAMA_API_URL',
+  OCO_BACKEND_ENDPOINT = 'OCO_BACKEND_ENDPOINT',
+  OCO_BACKEND_PATH = 'OCO_BACKEND_PATH',
   OCO_FLOWISE_ENDPOINT = 'OCO_FLOWISE_ENDPOINT',
   OCO_FLOWISE_API_KEY = 'OCO_FLOWISE_API_KEY'
 }
@@ -45,6 +47,7 @@ export enum CONFIG_MODES {
 
 export const MODEL_LIST = {
   openai: [
+    'gpt-4o-mini',
     'gpt-4o-mini',
     'gpt-3.5-turbo',
     'gpt-3.5-turbo-instruct',
@@ -132,8 +135,9 @@ export const configValidators = {
         config.OCO_ANTHROPIC_API_KEY ||
         config.OCO_AI_PROVIDER.startsWith('ollama') ||
         config.OCO_AZURE_API_KEY ||
-        config.OCO_AI_PROVIDER == 'test' ||
-        config.OCO_AI_PROVIDER == 'flowise',
+        config.OCO_AI_PROVIDER == 'flowise' ||
+        config.OCO_AI_PROVIDER == 'llmservice' ||
+        config.OCO_AI_PROVIDER == 'test',
       '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(
@@ -152,8 +156,9 @@ export const configValidators = {
         config.OCO_OPENAI_API_KEY ||
         config.OCO_AZURE_API_KEY ||
         config.OCO_AI_PROVIDER == 'ollama' ||
-        config.OCO_AI_PROVIDER == 'test' ||
-        config.OCO_AI_PROVIDER == 'flowise',
+        config.OCO_AI_PROVIDER == 'llmservice' ||
+        config.OCO_AI_PROVIDER == 'flowise' ||
+        config.OCO_AI_PROVIDER == 'test',
       'You need to provide an OpenAI/Anthropic/Azure API key'
     );
 
@@ -179,8 +184,9 @@ export const configValidators = {
       value ||
         config.OCO_OPENAI_API_KEY ||
         config.OCO_AI_PROVIDER == 'ollama' ||
-        config.OCO_AI_PROVIDER == 'test' ||
-        config.OCO_AI_PROVIDER == 'flowise',
+        config.OCO_AI_PROVIDER == 'llmservice' ||
+        config.OCO_AI_PROVIDER == 'flowise' ||
+        config.OCO_AI_PROVIDER == 'test',
       'You need to provide an OpenAI/Anthropic API key'
     );
 
@@ -323,9 +329,10 @@ export const configValidators = {
         'gemini',
         'azure',
         'test',
-        'flowise'
+        'flowise',
+        'llmservice'
       ].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', 'anthropic', 'azure', 'gemini', 'flowise', 'llmservice' or 'openai' (default)`
     );
     return value;
   },
@@ -379,7 +386,26 @@ export const configValidators = {
       `${value} is not a valid URL`
     );
     return value;
+  },
+
+  [CONFIG_KEYS.OCO_BACKEND_ENDPOINT](value: any) {
+    validateConfig(
+      CONFIG_KEYS.OCO_BACKEND_ENDPOINT,
+      typeof value === 'string',
+      'Must be string'
+    );
+    return value;
+  },
+
+  [CONFIG_KEYS.OCO_BACKEND_PATH](value: any) {
+    validateConfig(
+      CONFIG_KEYS.OCO_BACKEND_PATH,
+      typeof value === 'string',
+      'Must be string'
+    );
+    return value;
   }
+
 };
 
 export type ConfigType = {
@@ -423,6 +449,8 @@ export const getConfig = ({
       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_BACKEND_ENDPOINT: process.env.OCO_BACKEND_ENDPOINT || 'localhost:8000',
+    OCO_BACKEND_PATH: process.env.OCO_BACKEND_PATH || 'api/generate',
     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
@@ -452,6 +480,7 @@ export const getConfig = ({
         `Manually fix the '.env' file or global '~/.opencommit' config file.`
       );
 
+
       process.exit(1);
     }
   }

src/engine/llmservice.ts

@@ -0,0 +1,37 @@
+import axios, { AxiosError } from 'axios';
+import { ChatCompletionRequestMessage } from 'openai';
+import { AiEngine } from './Engine';
+import {
+  getConfig
+} from '../commands/config';
+
+const config = getConfig();
+
+
+export class LlmService implements AiEngine {
+
+  async generateCommitMessage(
+    messages: Array<ChatCompletionRequestMessage>
+  ): Promise<string | undefined> {
+
+    const gitDiff = messages[ messages.length - 1 ]?.content;
+    const url = `http://${config?.OCO_BACKEND_ENDPOINT}/${config?.OCO_BACKEND_PATH}`; 
+    const payload = {
+        user_prompt: gitDiff
+    }
+
+    try {
+      const response = await axios.post(url, payload, {
+        headers: {
+          'Content-Type': 'application/json'
+        }
+      });
+      const message = response.data;
+
+      return message;
+    } catch (err: any) {
+      const message = err.response?.data?.error ?? err.message;
+      throw new Error('local model issues. details: ' + message);
+    }
+  }
+}

src/utils/engine.ts

@@ -6,6 +6,7 @@ import { OllamaAi } from '../engine/ollama';
 import { AnthropicAi } from '../engine/anthropic'
 import { TestAi } from '../engine/testAi';
 import { Azure } from '../engine/azure';
+import { LlmService } from '../engine/llmservice';
 import { FlowiseAi } from '../engine/flowise'
 
 export function getEngine(): AiEngine {
@@ -28,6 +29,8 @@ export function getEngine(): AiEngine {
     return new Gemini();  
   } else if (provider == 'azure') {
   	return new Azure();
+  } else if(provider == 'llmservice'){
+    return new LlmService();
   } else if( provider == 'flowise'){
     return new FlowiseAi();
   }