Skip to content

Commit 24b38b6

Browse files
milos85vasicclaude
milos85vasic
and
claude
committed
Add comprehensive unit tests, documentation, and fallback visualization
Unit Tests: - Add ensemble_test.go with 18 tests for LLM ensemble functionality - Add background_task_handler_test.go with 37+ tests for task management - Add redis_test.go with 32 tests using miniredis for cache operations - Add task_queue_test.go with 25 tests for PostgresTaskQueue - Add debate_visualization.go/test for multi-pass debate formatting Package Documentation (10 READMEs): - internal/background: Task queue, worker pool, resource monitoring - internal/notifications: SSE, WebSocket, Webhooks, Polling - internal/llm: Provider interface, circuit breaker, health monitor - internal/services: Provider registry, debate system, protocols - internal/handlers: HTTP handlers, OpenAI compatibility - internal/optimization: GPTCache, streaming, pipeline - internal/verifier: Startup verification, scoring - internal/agents: CLI agent registry (18 agents) - internal/tools: Tool schema registry (21 tools) - internal/plugins: Hot-reloadable plugin system Website Content: - Add features.html page with provider grid - Add 07-protocols.md user manual (MCP/LSP/ACP) - Add 08-troubleshooting.md user manual - Add course-05-protocols.md video course - Add course-06-testing.md video course Fallback Chain Visualization: - Add FormatFallbackChainIndicator for multi-chain fallbacks - Add FormatFallbackChainWithContent with response content - Add formatFallbackReason for user-friendly error messages - Use darker timing colors (ANSIDim+ANSIBrightBlack) - Support multiple chained fallbacks with individual timing - Example: [A: Sonnet 4.5] <--- (10 ms) [Fallback, Rate limit reached: DeepSeek] (100 ms) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
1 parent 60d74ab commit 24b38b6

26 files changed

+9242
-81
lines changed

‎COMPREHENSIVE_COMPLETION_PLAN.md‎

Lines changed: 657 additions & 0 deletions
Large diffs are not rendered by default.

‎Website/public/features.html‎

Lines changed: 216 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,216 @@
1+
<!DOCTYPE html>
2+
<html lang="en">
3+
<head>
4+
<meta charset="UTF-8">
5+
<meta name="viewport" content="width=device-width, initial-scale=1.0">
6+
<title>Features - HelixAgent</title>
7+
<link rel="stylesheet" href="styles/main.css">
8+
<meta name="description" content="HelixAgent features - AI ensemble, multi-provider support, AI debate system, and more">
9+
</head>
10+
<body>
11+
<header>
12+
<nav>
13+
<a href="/" class="logo">HelixAgent</a>
14+
<ul>
15+
<li><a href="features.html" class="active">Features</a></li>
16+
<li><a href="pricing.html">Pricing</a></li>
17+
<li><a href="docs/">Documentation</a></li>
18+
<li><a href="contact.html">Contact</a></li>
19+
</ul>
20+
</nav>
21+
</header>
22+
23+
<main>
24+
<section class="hero">
25+
<h1>Powerful AI Ensemble Features</h1>
26+
<p>Combine the best LLMs into one unified API with intelligent orchestration</p>
27+
</section>
28+
29+
<section class="features-grid">
30+
<article class="feature-card">
31+
<h2>Multi-Provider Ensemble</h2>
32+
<p>Query multiple LLM providers simultaneously and get the best response through intelligent aggregation.</p>
33+
<ul>
34+
<li>18+ LLM providers supported</li>
35+
<li>Automatic provider selection</li>
36+
<li>Confidence-weighted voting</li>
37+
<li>Fallback chain management</li>
38+
</ul>
39+
</article>
40+
41+
<article class="feature-card">
42+
<h2>AI Debate System</h2>
43+
<p>Let AI models debate and reach consensus on complex topics for more reliable answers.</p>
44+
<ul>
45+
<li>Multi-round debates</li>
46+
<li>5 positions x 3 LLMs = 15 perspectives</li>
47+
<li>Multi-pass validation</li>
48+
<li>Consensus synthesis</li>
49+
</ul>
50+
</article>
51+
52+
<article class="feature-card">
53+
<h2>Dynamic Provider Selection</h2>
54+
<p>Providers are automatically ranked and selected based on real-time verification scores.</p>
55+
<ul>
56+
<li>5-component scoring algorithm</li>
57+
<li>Response speed (25%)</li>
58+
<li>Cost effectiveness (25%)</li>
59+
<li>Model capability (20%)</li>
60+
</ul>
61+
</article>
62+
63+
<article class="feature-card">
64+
<h2>OpenAI-Compatible API</h2>
65+
<p>Drop-in replacement for OpenAI API with enhanced multi-provider capabilities.</p>
66+
<ul>
67+
<li>Chat completions endpoint</li>
68+
<li>Streaming support</li>
69+
<li>Tool/function calling</li>
70+
<li>Embeddings generation</li>
71+
</ul>
72+
</article>
73+
74+
<article class="feature-card">
75+
<h2>Protocol Support</h2>
76+
<p>Integrate with development tools through standardized protocols.</p>
77+
<ul>
78+
<li>MCP (Model Context Protocol)</li>
79+
<li>LSP (Language Server Protocol)</li>
80+
<li>ACP (Agent Communication Protocol)</li>
81+
<li>Real-time SSE/WebSocket</li>
82+
</ul>
83+
</article>
84+
85+
<article class="feature-card">
86+
<h2>Background Task Execution</h2>
87+
<p>Run long-running tasks asynchronously with progress tracking and notifications.</p>
88+
<ul>
89+
<li>Task queue with priorities</li>
90+
<li>Worker pool auto-scaling</li>
91+
<li>Pause/resume capability</li>
92+
<li>Resource monitoring</li>
93+
</ul>
94+
</article>
95+
96+
<article class="feature-card">
97+
<h2>Intelligent Caching</h2>
98+
<p>Reduce costs and latency with semantic caching of LLM responses.</p>
99+
<ul>
100+
<li>GPT Cache integration</li>
101+
<li>Redis-backed storage</li>
102+
<li>Similarity-based matching</li>
103+
<li>Configurable TTL</li>
104+
</ul>
105+
</article>
106+
107+
<article class="feature-card">
108+
<h2>Knowledge Graph (Cognee)</h2>
109+
<p>Enhance responses with knowledge graph and RAG capabilities.</p>
110+
<ul>
111+
<li>Automatic knowledge extraction</li>
112+
<li>Graph-based retrieval</li>
113+
<li>Context augmentation</li>
114+
<li>Memory persistence</li>
115+
</ul>
116+
</article>
117+
118+
<article class="feature-card">
119+
<h2>Fault Tolerance</h2>
120+
<p>Built-in resilience for production deployments.</p>
121+
<ul>
122+
<li>Circuit breaker pattern</li>
123+
<li>Automatic retry with backoff</li>
124+
<li>Provider health monitoring</li>
125+
<li>Graceful degradation</li>
126+
</ul>
127+
</article>
128+
129+
<article class="feature-card">
130+
<h2>Plugin System</h2>
131+
<p>Extend functionality with hot-reloadable plugins.</p>
132+
<ul>
133+
<li>Hot reload without restart</li>
134+
<li>Dependency resolution</li>
135+
<li>Security sandboxing</li>
136+
<li>Version management</li>
137+
</ul>
138+
</article>
139+
140+
<article class="feature-card">
141+
<h2>Semantic Intent Detection</h2>
142+
<p>Understand user intent through AI-powered semantic analysis.</p>
143+
<ul>
144+
<li>Zero hardcoded patterns</li>
145+
<li>LLM-based classification</li>
146+
<li>Context-aware understanding</li>
147+
<li>Multi-language support</li>
148+
</ul>
149+
</article>
150+
151+
<article class="feature-card">
152+
<h2>Enterprise Security</h2>
153+
<p>Production-ready security features for enterprise deployments.</p>
154+
<ul>
155+
<li>API key authentication</li>
156+
<li>OAuth 2.0 support</li>
157+
<li>Rate limiting</li>
158+
<li>Audit logging</li>
159+
</ul>
160+
</article>
161+
</section>
162+
163+
<section class="providers">
164+
<h2>Supported Providers</h2>
165+
<div class="provider-grid">
166+
<div class="provider"><img src="assets/images/providers/claude.svg" alt="Claude"><span>Claude</span></div>
167+
<div class="provider"><img src="assets/images/providers/deepseek.svg" alt="DeepSeek"><span>DeepSeek</span></div>
168+
<div class="provider"><img src="assets/images/providers/gemini.svg" alt="Gemini"><span>Gemini</span></div>
169+
<div class="provider"><img src="assets/images/providers/qwen.svg" alt="Qwen"><span>Qwen</span></div>
170+
<div class="provider"><img src="assets/images/providers/openrouter.svg" alt="OpenRouter"><span>OpenRouter</span></div>
171+
<div class="provider"><img src="assets/images/providers/ollama.svg" alt="Ollama"><span>Ollama</span></div>
172+
<div class="provider"><img src="assets/images/providers/zai.svg" alt="ZAI"><span>ZAI</span></div>
173+
<div class="provider">Mistral</div>
174+
<div class="provider">Cerebras</div>
175+
<div class="provider">Zen</div>
176+
</div>
177+
</section>
178+
179+
<section class="cta">
180+
<h2>Ready to Get Started?</h2>
181+
<p>Deploy HelixAgent and unlock the power of AI ensemble in minutes.</p>
182+
<a href="docs/" class="btn-primary">View Documentation</a>
183+
<a href="https://github.com/helix-agent/helixagent" class="btn-secondary">GitHub Repository</a>
184+
</section>
185+
</main>
186+
187+
<footer>
188+
<div class="footer-content">
189+
<div class="footer-section">
190+
<h3>HelixAgent</h3>
191+
<p>AI-powered ensemble LLM service</p>
192+
</div>
193+
<div class="footer-section">
194+
<h3>Links</h3>
195+
<ul>
196+
<li><a href="features.html">Features</a></li>
197+
<li><a href="pricing.html">Pricing</a></li>
198+
<li><a href="docs/">Documentation</a></li>
199+
</ul>
200+
</div>
201+
<div class="footer-section">
202+
<h3>Legal</h3>
203+
<ul>
204+
<li><a href="privacy.html">Privacy Policy</a></li>
205+
<li><a href="terms.html">Terms of Service</a></li>
206+
</ul>
207+
</div>
208+
</div>
209+
<div class="footer-bottom">
210+
<p>&copy; 2024 HelixAgent. All rights reserved.</p>
211+
</div>
212+
</footer>
213+
214+
<script src="scripts/main.js"></script>
215+
</body>
216+
</html>

0 commit comments

Comments
 (0)