[Plugin] AHDA LITE - Actually helpful digital assistant - control PC …

…with OMI (#1116)

See https://github.com/ActuallyAdvanced/OMI-AHDA

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

- **New Features**
- Introduced two new plugins: "Gen Z/A Translator" for translating
sentences into Gen Z or Gen Alpha slang.
- Added the "AHDA" plugin for voice control of your PC, enhancing user
accessibility.

- **Documentation**
	- Updated README with a new section for AHDA setup instructions.

- **Chores**
- Implemented webhooks and configuration endpoints for AHDA integration.
	- Added utility functions for managing AHDA-related configurations.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

community-plugins.json

@@ -653,7 +653,34 @@
     "capabilities": [
         "memories"
     ],
-    "memory_prompt": "Please transform the following sentence into Gen Z/Gen Alpha slang, incorporating as many of the following terms as possible (and more): *bet, fax, W, vibe, bussin', rizz, lit, drip, mid, cap, salty, slay, extra, flex, yeet, mood, glow up, big L, bop, Gucci, period, sus, sheesh, lowkey, snack, snatched, stan, thirsty, tea, ghost, cringe, delulu, fire, basic, GOAT, ick, peep, sleeper, low-vibrational, it's giving,* and *vibin'*, cooked. The goal is to maintain the original meaning of the sentence but make it sound as if it's being spoken by a Gen Z or Gen Alpha person. Feel free to get creative with the slang, keeping it short and casual, while giving the sentence that trendy, social-media-ready vibe. Also, ensure the use of terms fits contextually and naturally into the sentence. Only respond with the translated sentance.",
+    "memory_prompt": "Please transform the following sentence into Gen Z/Gen Alpha slang, incorporating as many of the following terms as possible (and more): *bet, fax, W, vibe, bussin', rizz, lit, drip, mid, cap, salty, slay, extra, flex, yeet, mood, glow up, big L, bop, Gucci, period, sus, sheesh, lowkey, snack, snatched, stan, thirsty, tea, ghost, cringe, delulu, fire, basic, GOAT, ick, peep, sleeper, low-vibrational, it's giving,* and *vibin'*, cooked. The goal is to maintain the original meaning of the sentence but make it sound as if it's being spoken by a Gen Z or Gen Alpha person. Feel free to get creative with the slang, keeping it short and casual, while giving the sentence that trendy, social-media-ready vibe. Also, ensure the use of terms fits contextually and naturally into the sentence.",
+    "deleted": false
+},
+  {
+    "id": "ahda",
+    "name": "AHDA",
+    "author": "Neo",
+    "description": "Control your PC with your voice",
+    "image": "/plugins/logos/AHDA.png",
+    "capabilities": [
+      "external_integration"
+    ],
+    "external_integration": {
+      "triggers_on": "transcript_processed",
+      "webhook_url": "https://based-hardware--plugins-api.modal.run/ahda/send-webhook",
+      "setup_completed_url": null,
+      "setup_instructions_file_path": "/plugins/instructions/ahda/README.md",
+      "auth_steps": [
+        {
+          "name": "Set up and install AHDA on your PC",
+          "url": "https://github.com/ActuallyAdvanced/OMI-AHDA/blob/main/README.md"
+        },
+        {
+          "name": "Enter URL into OMI",
+          "url": "https://based-hardware--plugins-api.modal.run/ahda/index"
+        }
+      ]
+    },
     "deleted": false
 }
 ]

plugins/example/ahda/client.py

@@ -0,0 +1,143 @@
+from fastapi import APIRouter, Request, HTTPException, Form, Query, BackgroundTasks, Body
+from fastapi.responses import HTMLResponse, FileResponse, JSONResponse
+from db import get_ahda_url, store_ahda, get_ahda_os
+import os
+import requests
+from models import RealtimePluginRequest, EndpointResponse
+import time
+import asyncio
+import logging
+from langchain_openai import ChatOpenAI
+
+router = APIRouter()
+
+active_sessions = {}
+
+KEYWORD = "computer"
+COMMAND_TIMEOUT = 5  # Seconds to wait after the last word to finalize the command
+
+# Path to the directory containing `index.html`
+BASE_DIR = os.path.dirname(os.path.abspath(__file__))
+INDEX_PATH = os.path.join(BASE_DIR, "index.html")
+
+chat = ChatOpenAI(model='gpt-4o', temperature=0)
+
+# Use requests to get raw text from URL
+prompt = requests.get("https://raw.githubusercontent.com/ActuallyAdvanced/OMI-AHDA/main/prompt.txt").text
+
+# Configure logging
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+# AHDA Utils
+def sendToPC(uid, response):
+    ahda_url = get_ahda_url(uid)
+    if not ahda_url:
+        raise ValueError('AHDA URL not configured for this UID')
+    payload = {
+        'uid': uid,
+        'response': response
+    }
+    try:
+        resp = requests.post(ahda_url+"/recieve", json=payload)
+        resp.raise_for_status()
+    except requests.RequestException as e:
+        logger.error(f"Error sending webhook: {e}")
+        raise
+    return {'message': 'Webhook sent successfully'}
+
+
+@router.post('/ahda/send-webhook', tags=['ahda', 'realtime'])
+async def send_ahda_webhook(
+    uid: str = Query(...), 
+    data: dict = Body(...),
+    background_tasks: BackgroundTasks = BackgroundTasks()
+):
+    segments = data.get("segments")
+    if not uid:
+        raise HTTPException(status_code=400, detail="UID is required")
+
+    if not segments or not isinstance(segments, list):
+        raise HTTPException(status_code=400, detail="Invalid payload")
+
+    if uid not in active_sessions:
+        logger.info(f"New session started: {uid}")
+        active_sessions[uid] = {
+            "command": "",
+            "last_received_time": time.time(),
+            "active": False,
+            "timer": None
+        }
+
+    async def schedule_finalize_command(uid, delay):
+        await asyncio.sleep(delay)
+        await finalize_command(uid)
+
+    async def finalize_command(uid):
+        final_command = active_sessions[uid]["command"].strip()
+        if final_command:
+            logger.info(f"Final command for session {uid}: {final_command}")
+            await call_chatgpt_to_generate_code(final_command, uid)
+        active_sessions[uid]["command"] = ""
+        active_sessions[uid]["active"] = False
+        active_sessions[uid]["timer"] = None
+
+    # Adjusted to handle segments as dictionaries
+    for segment in segments:
+        text = segment.get("text", "").strip().lower()
+        logger.info(f"Received segment: {text} (session_id: {uid})")
+
+        if KEYWORD in text:
+            logger.info("Activation keyword detected!")
+            active_sessions[uid]["active"] = True
+            active_sessions[uid]["last_received_time"] = time.time()
+
+            if active_sessions[uid]["timer"]:
+                pass
+
+            active_sessions[uid]["timer"] = background_tasks.add_task(
+                schedule_finalize_command, uid, COMMAND_TIMEOUT
+            )
+            continue
+
+        if active_sessions[uid]["active"]:
+            active_sessions[uid]["command"] += " " + text
+            active_sessions[uid]["last_received_time"] = time.time()
+            logger.info(f"Aggregating command: {active_sessions[uid]['command'].strip()}")
+
+            if active_sessions[uid]["timer"]:
+                pass
+
+            active_sessions[uid]["timer"] = background_tasks.add_task(
+                schedule_finalize_command, uid, COMMAND_TIMEOUT
+            )
+
+    return {"status": "success"}
+
+
+async def call_chatgpt_to_generate_code(command, uid):
+    try:
+        ahda_os = get_ahda_os(uid)
+        messages = [
+            ("system", prompt.replace("{os_name}",ahda_os)),
+            ("human", command),
+        ]
+        ai_msg = chat.invoke(messages)
+        sendToPC(uid, ai_msg)
+    except Exception as e:
+        logger.error(f"Error calling ChatGPT-4: {e}")
+        return {"type": "error", "content": str(e)}
+
+@router.get('/ahda/index', response_class=HTMLResponse, tags=['ahda'])
+async def get_ahda_index(request: Request, uid: str = Query(None)):
+    if not uid:
+        raise HTTPException(status_code=400, detail="UID is required")
+    return FileResponse(INDEX_PATH)
+
+@router.post('/ahda/configure', tags=['ahda'])
+def configure_ahda(uid: str = Form(...), url: str = Form(...), os: str = Form(...)):
+    if not uid or not url:
+        raise HTTPException(status_code=400, detail="Both UID, URL AND OS are required")
+
+    store_ahda(uid, url, os)
+    return {'message': 'AHDA configured successfully'}

plugins/example/ahda/index.html

@@ -0,0 +1,138 @@
+<!DOCTYPE html>
+<html lang="en">
+
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>AHDA Integration</title>
+    <link rel="preconnect" href="https://fonts.googleapis.com">
+    <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
+    <link href="https://fonts.googleapis.com/css2?family=Assistant:[email protected]&display=swap" rel="stylesheet">
+    <style>
+        body {
+            font-family: Assistant, sans-serif;
+            background-color: #f4f4f4;
+            margin: 0;
+            padding: 0;
+            display: flex;
+            justify-content: center;
+            align-items: center;
+            height: 100vh;
+        }
+
+        .container {
+            background-color: #fff;
+            padding: 3rem;
+            border-radius: 8px;
+            box-shadow: 0 0 10px rgba(0, 0, 0, 0.1);
+            text-align: center;
+            max-width: 400px;
+            width: 100%;
+        }
+
+        h1 {
+            margin-bottom: 2rem;
+        }
+
+        .logo-container {
+            display: flex;
+            justify-content: center;
+            gap: 20px;
+            margin-bottom: 2rem;
+        }
+
+        .logo-container img {
+            width: 80px;
+            height: 80px;
+            object-fit: contain;
+        }
+
+        input[type="text"],
+        button {
+            width: 100%;
+            padding: 10px;
+            margin-bottom: 1rem;
+            border: 1px solid #ddd;
+            border-radius: 4px;
+            box-sizing: border-box;
+        }
+
+        button {
+            background-color: #000;
+            color: #fff;
+            border: none;
+            cursor: pointer;
+            font-size: 1rem;
+        }
+
+        button:hover {
+            background-color: #333;
+        }
+
+        p {
+            margin-top: 1rem;
+        }
+
+        .plugin-text {
+            font-size: 1.2rem;
+            font-weight: bold;
+            margin-bottom: 1rem;
+        }
+    </style>
+</head>
+
+<body>
+    <div class="container">
+        <div class="logo-container">
+            <img src="https://avatars.githubusercontent.com/u/162546372" alt="OMI Logo" />
+            <img src="https://avatars.githubusercontent.com/u/181646370" alt="AHDA Logo" />
+        </div>
+        <p class="plugin-text">AHDA Integration Plugin</p>
+        <h1>AHDA Integration</h1>
+        <h1>Install https://github.com/ActuallyAdvanced/OMI-AHDA first</h1>
+        <form id="ahda-form">
+            <input type="text" id="url" placeholder="Enter AHDA URL" required>
+            <input type="text" id="os" placeholder="Enter Operating System" required>
+            <button type="submit">Save Configuration</button>
+        </form>
+        <p id="response-message"></p>
+    </div>
+
+    <script>
+        // Extract the uid from the URL query parameters
+        const urlParams = new URLSearchParams(window.location.search);
+        const uid = urlParams.get('uid');
+
+        if (!uid) {
+            document.getElementById('response-message').textContent = "UID is missing. Please check the URL.";
+            document.getElementById('response-message').style.color = 'red';
+        }
+
+        document.getElementById('ahda-form').addEventListener('submit', async function (event) {
+            event.preventDefault();
+
+            const url = document.getElementById('url').value;
+            const os = document.getElementById('os').value;
+
+            const response = await fetch('/ahda/configure', {
+                method: 'POST',
+                headers: {
+                    'Content-Type': 'application/x-www-form-urlencoded',
+                },
+                body: new URLSearchParams({ uid, url, os })
+            });
+
+            const result = await response.json();
+            const messageElement = document.getElementById('response-message');
+            if (response.ok) {
+                messageElement.textContent = result.message;
+                messageElement.style.color = 'green';
+            } else {
+                messageElement.textContent = result.detail || 'An error occurred. Please try again.';
+                messageElement.style.color = 'red';
+            }
+        });
+    </script>
+</body>
+
+</html>

plugins/example/db.py

@@ -132,3 +132,19 @@ def set_task_result(task_id: str, result: str):
 def get_task_result(task_id: str) -> str:
     result = r.get(f'task_result:{task_id}')
     return result.decode('utf-8') if result else None
+
+# **********************************************************
+# ************ AHDA UTILS (PC Control) ************
+# **********************************************************
+
+def store_ahda(uid: str, url: str, os: str):
+    r.set(f'ahda_url:{uid}', url)
+    r.set(f'ahda_os:{uid}', os)
+
+def get_ahda_url(uid: str) -> str:
+    val = r.get(f'ahda_url:{uid}')
+    return val.decode('utf-8') if val else None
+
+def get_ahda_os(uid: str) -> str:
+    val = r.get(f'ahda_os:{uid}')
+    return val.decode('utf-8') if val else None

plugins/instructions/ahda/README.md

@@ -0,0 +1 @@
+### [Setup Guide for AHDA](https://github.com/ActuallyAdvanced/OMI-AHDA/blob/main/README.md)