diff --git a/AGENT_GUIDE.md b/AGENT_GUIDE.md index ddd2dad..74f9dd5 100644 --- a/AGENT_GUIDE.md +++ b/AGENT_GUIDE.md @@ -97,3 +97,31 @@ Use get_context() to see their motivation profile: understand it, but don't lecture. - Don't read back integration credentials. You can check status but never retrieve stored secrets. + +## Multi-Device Integration + +If you have access to both Diligence tools and a device MCP (like COROS), +you can bridge data between them: + +1. Read the user's recent workouts from the device MCP +2. Use Diligence's log_activity tool to import them +3. This earns the user points automatically + +Example workflow with COROS MCP connected: +- Check COROS for today's activities +- For each unlogged activity, call log_activity with the details +- Report the points earned + +## Strava Data Warning + +Strava's API terms prohibit use of their data for AI/ML purposes. +If the user has Strava connected, you may see synced activities in +their log, but do not reference Strava-specific data in your coaching +advice. Treat Strava activities as user-reported workouts only. + +## Provider-Agnostic + +This guide works regardless of which LLM powers the coaching. +The same context, tools, and personality apply whether you are +GPT-4o, Claude, Llama, Gemini, or any other model. The user +chose you — respect their choice and focus on being helpful. diff --git a/README.md b/README.md index f080c3f..37fb408 100644 --- a/README.md +++ b/README.md @@ -108,34 +108,65 @@ You should see 4 containers, all healthy: `frontend`, `backend`, `mcp-connector` - **Program tracking** — 90-day structured programs (StrongLifts, Darebee, etc.) with day-by-day progression. - **Configurable rewards** — you define what's worth earning. Gaming time, screen time, treats — your rules. -## Connecting an AI Agent +## Built-in AI Coach -Point your agent's MCP config at: +Diligence includes a built-in AI coaching chat. Configure any LLM provider in **Settings → Integrations**, then open the **Coach** tab. -| Environment | URL | -|-------------|-----| -| Local (development) | `http://localhost:3001/sse` | -| Behind reverse proxy | `https://your-domain/mcp` | +Supported providers (one API key, that's it): -Works with Claude Desktop, Cursor, Windsurf, and any MCP-compatible agent. +| Provider | Free tier | What you get | +|----------|-----------|-------------| +| **OpenRouter** | 26 free models | 300+ models from every major provider, one key | +| **Hugging Face** | Yes | Thousands of open-source models | +| **Groq** | Yes | Ultra-fast Llama inference | +| **Ollama** | Local, free | Run any model on your own machine | +| **OpenAI** | No | GPT-4o, GPT-4o-mini | +| **Claude** | No | Claude Sonnet 4.6, Opus 4.8 | +| **Gemini** | Yes | Gemini 2.0 Flash, 2.5 Pro | +| **Custom** | — | Any OpenAI-compatible endpoint (vLLM, TGI, LiteLLM) | -Copy the contents of [AGENT_GUIDE.md](AGENT_GUIDE.md) into your agent's system instructions for motivation-aware coaching. +The AI coach has access to your profile, points, program schedule, and motivation type. It can log workouts, search food, and create meal plans through the chat. -### Claude Desktop example +### Connecting external AI agents (MCP) -Add to your `claude_desktop_config.json`: +The MCP connector at port 3001 works with any MCP-compatible agent. The built-in chat and external agents coexist — use whichever fits your workflow. +**Claude Desktop** — add to `claude_desktop_config.json`: ```json { "mcpServers": { - "diligence": { - "url": "http://localhost:3001/sse" - } + "diligence": { "url": "http://localhost:3001/sse" } } } ``` -Then paste the contents of `AGENT_GUIDE.md` into your Claude Desktop project instructions. +**Claude Code** (CLI): +```bash +claude mcp add diligence --transport sse http://localhost:3001/sse +``` + +**Cursor** — add to `.cursor/mcp.json`: +```json +{ + "mcpServers": { + "diligence": { "url": "http://localhost:3001/sse" } + } +} +``` + +**Windsurf** — add to MCP settings, same format as Cursor. + +**COROS watch owners** — add both Diligence and COROS MCP servers to your agent. The agent bridges your watch data with your fitness log: +```json +{ + "mcpServers": { + "diligence": { "url": "http://localhost:3001/sse" }, + "coros": { "url": "https://your-coros-mcp-url/sse" } + } +} +``` + +Copy the contents of [AGENT_GUIDE.md](AGENT_GUIDE.md) into your agent's system instructions for motivation-aware coaching. ## Configuring Integrations diff --git a/frontend/src/App.jsx b/frontend/src/App.jsx index 717b218..e69de29 100644 --- a/frontend/src/App.jsx +++ b/frontend/src/App.jsx @@ -1,128 +0,0 @@ -import { Routes, Route, Navigate, NavLink, useNavigate, useLocation } from 'react-router-dom' -import { useState, useEffect } from 'react' -import { hasToken, clearToken, api } from './api' -import Login from './pages/Login' -import Dashboard from './pages/Dashboard' -import LogActivity from './pages/LogActivity' -import LogFood from './pages/LogFood' -import Nutrition from './pages/Nutrition' -import Rewards from './pages/Rewards' -import Settings from './pages/Settings' -import Onboarding from './pages/Onboarding' -import WeekView from './pages/WeekView' -import Welcome from './pages/Welcome' -import SettingsIntegrations from './pages/SettingsIntegrations' -import MealPlan from './pages/MealPlan' -import ProgramSearch from './pages/ProgramSearch' -import ProgramDetail from './pages/ProgramDetail' -import CatalogDetail from './pages/CatalogDetail' -import Support from './pages/Support' -import SupportAdmin from './pages/SupportAdmin' - -function ProtectedRoute({ children }) { - if (!hasToken()) return - return children -} - -function HelpButton() { - const [unread, setUnread] = useState(0) - const navigate = useNavigate() - const location = useLocation() - - // Don't show on login/onboarding/support pages - const hidden = ['/login', '/onboarding', '/support'].some(p => location.pathname.startsWith(p)) - - useEffect(() => { - if (hidden || !hasToken()) return - api.getUnreadCount() - .then(d => setUnread(d.unread || 0)) - .catch(() => {}) - // Poll every 60s for new replies - const interval = setInterval(() => { - api.getUnreadCount() - .then(d => setUnread(d.unread || 0)) - .catch(() => {}) - }, 60000) - return () => clearInterval(interval) - }, [location.pathname, hidden]) - - if (hidden) return null - - return ( - - ) -} - -function NavBar() { - return ( - - ) -} - -export default function App() { - return ( - <> - {hasToken() && } - - } /> - } /> - } /> - } /> - } /> - } /> - } /> - } /> - } /> - } /> - } /> - } /> - } /> - } /> - } /> - } /> - } /> - } /> - - {hasToken() && } - - ) -} diff --git a/frontend/src/api.js b/frontend/src/api.js index 93db0f1..c027719 100644 --- a/frontend/src/api.js +++ b/frontend/src/api.js @@ -139,6 +139,10 @@ export const api = { listProviders: () => request('/integrations/providers'), configureIntegration: (provider, credentials) => request('/integrations/configure', { method: 'POST', body: JSON.stringify({ provider, credentials }) }), + // AI Coaching + getAIStatus: () => request('/ai/status'), + // Note: chatWithAI uses fetch directly in ChatCoach.jsx for SSE streaming + // Resources getResourceRecommendations: () => request('/onboarding/recommendations'), }; diff --git a/frontend/src/pages/ChatCoach.jsx b/frontend/src/pages/ChatCoach.jsx new file mode 100644 index 0000000..abfcdcf --- /dev/null +++ b/frontend/src/pages/ChatCoach.jsx @@ -0,0 +1,194 @@ +import { useState, useEffect, useRef } from 'react' +import { api } from '../api' + +export default function ChatCoach() { + const [messages, setMessages] = useState([]) + const [input, setInput] = useState('') + const [streaming, setStreaming] = useState(false) + const [aiStatus, setAiStatus] = useState(null) + const scrollRef = useRef(null) + + useEffect(() => { + api.getAIStatus().then(setAiStatus).catch(() => setAiStatus({ configured: false })) + }, []) + + useEffect(() => { + scrollRef.current?.scrollIntoView({ behavior: 'smooth' }) + }, [messages]) + + const sendMessage = async () => { + const text = input.trim() + if (!text || streaming) return + + const userMsg = { role: 'user', content: text } + setMessages(prev => [...prev, userMsg]) + setInput('') + setStreaming(true) + + // Add placeholder for assistant + setMessages(prev => [...prev, { role: 'assistant', content: '' }]) + + try { + const history = messages.map(m => ({ role: m.role, content: m.content })) + const token = localStorage.getItem('fitness_token') + const resp = await fetch('/api/ai/chat', { + method: 'POST', + headers: { + 'Content-Type': 'application/json', + ...(token ? { Authorization: `Bearer ${token}` } : {}), + }, + body: JSON.stringify({ message: text, history }), + }) + + const reader = resp.body.getReader() + const decoder = new TextDecoder() + let buffer = '' + + while (true) { + const { done, value } = await reader.read() + if (done) break + + buffer += decoder.decode(value, { stream: true }) + const lines = buffer.split('\n') + buffer = lines.pop() || '' + + for (const line of lines) { + if (line.startsWith('data: ') && line.trim() !== 'data: [DONE]') { + try { + const { text: chunk } = JSON.parse(line.slice(6)) + if (chunk) { + setMessages(prev => { + const updated = [...prev] + const last = updated[updated.length - 1] + updated[updated.length - 1] = { ...last, content: last.content + chunk } + return updated + }) + } + } catch {} + } + } + } + } catch (err) { + setMessages(prev => { + const updated = [...prev] + updated[updated.length - 1] = { + role: 'assistant', + content: 'Connection error. Check that the backend is running and an AI provider is configured in Settings → Integrations.', + } + return updated + }) + } + setStreaming(false) + } + + const suggestions = [ + 'What should I eat today?', + 'Log my 30-minute run', + 'How am I doing this week?', + 'Create a meal plan for me', + ] + + if (aiStatus && !aiStatus.configured) { + return ( +
+

AI Coach

+
+
🤖
+

No AI provider connected

+

+ Connect OpenAI, OpenRouter, Claude, Ollama, or any other provider to start chatting with your AI fitness coach. +

+ + Configure AI Provider + +
+
+ ) + } + + return ( +
+
+

AI Coach

+ {aiStatus?.provider && ( + + {aiStatus.provider} · {aiStatus.model} + + )} +
+ +
+ {messages.length === 0 && ( +
+

+ Ask me anything about your fitness, nutrition, or program. +

+
+ {suggestions.map(s => ( + + ))} +
+
+ )} + + {messages.map((msg, i) => ( +
+
+ {msg.content || (streaming && i === messages.length - 1 ? '...' : '')} +
+
+ ))} +
+
+ +
+ setInput(e.target.value)} + onKeyDown={e => e.key === 'Enter' && !e.shiftKey && sendMessage()} + placeholder={streaming ? 'Thinking...' : 'Ask your AI coach...'} + disabled={streaming} + style={{ flex: 1 }} + /> + +
+
+ ) +}