Skip to content

Commit b1f8df1

Browse files
author
ZENTAXA Team
committed
Update documentation with live platform URL and SDK reference
- Add live platform link: zentaxaapp.azurewebsites.net - Add SDK installation and usage examples - Add SDK Reference section with trace(), log(), metric() methods - Add multi-step agent visualization guide - Update module descriptions with correct URLs - Add PyPI package link - Update glossary with SDK terms
1 parent d4f39a8 commit b1f8df1

File tree

2 files changed

+182
-31
lines changed

2 files changed

+182
-31
lines changed

‎README.md‎

Lines changed: 7 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,12 @@
11
# ZENTAXA Landing Page
22

3-
A modern, responsive landing page for ZENTAXA - AI Agent Management Platform.
3+
A modern, responsive landing page for ZENTAXA - AI Agent Observability Platform.
4+
5+
## 🌐 Live Links
6+
7+
- **Platform:** [zentaxaapp.azurewebsites.net](https://zentaxaapp.azurewebsites.net)
8+
- **PyPI SDK:** [pypi.org/project/zentaxa](https://pypi.org/project/zentaxa/)
9+
- **Documentation:** [View Docs](documentation.html)
410

511
## 🚀 Features
612

‎documentation.html‎

Lines changed: 175 additions & 30 deletions
Original file line numberDiff line numberDiff line change
@@ -208,7 +208,15 @@ <h1 class="doc-title">User Documentation</h1>
208208
<li><a href="#troubleshooting">5. Troubleshooting</a></li>
209209
<li><a href="#best-practices">6. Best Practices</a></li>
210210
<li><a href="#glossary">7. Glossary</a></li>
211+
<li><a href="#sdk-reference">8. SDK Reference</a></li>
211212
</ul>
213+
<div style="margin-top: 1.5rem; padding-top: 1rem; border-top: 1px solid var(--border);">
214+
<p style="font-size: 0.8rem; color: var(--text-muted); margin-bottom: 0.5rem;">Quick Links</p>
215+
<ul style="font-size: 0.85rem;">
216+
<li><a href="https://zentaxaapp.azurewebsites.net" target="_blank">🚀 Live Platform</a></li>
217+
<li><a href="https://pypi.org/project/zentaxa/" target="_blank">📦 PyPI Package</a></li>
218+
</ul>
219+
</div>
212220
</nav>
213221
</aside>
214222

@@ -272,24 +280,65 @@ <h4>Multi-agent Observability</h4>
272280
<!-- 3. Getting Started -->
273281
<section id="getting-started">
274282
<h2>3. Getting Started</h2>
275-
<p>Follow these simple steps to start using the Zentaxa dashboard.</p>
283+
<p>Follow these simple steps to start using Zentaxa with your AI agents.</p>
284+
285+
<h3>Live Platform</h3>
286+
<p>Access the live Zentaxa platform at: <a href="https://zentaxaapp.azurewebsites.net" target="_blank" style="color: var(--primary);">https://zentaxaapp.azurewebsites.net</a></p>
287+
288+
<h3>Install the SDK</h3>
289+
<pre style="background: var(--surface); padding: 1rem; border-radius: var(--radius); border: 1px solid var(--border); overflow-x: auto;"><code>pip install zentaxa</code></pre>
290+
291+
<h3>Quick Integration</h3>
292+
<pre style="background: var(--surface); padding: 1rem; border-radius: var(--radius); border: 1px solid var(--border); overflow-x: auto; margin-bottom: 1.5rem;"><code>from zentaxa import ZentaxaClient
293+
294+
# Connects to live platform by default
295+
client = ZentaxaClient()
296+
297+
# Start tracking an agent run
298+
response = client.trace(
299+
run_id=None,
300+
agent_id="my-agent",
301+
framework="langchain", # or crewai, autogen, custom
302+
event_type="agent_start",
303+
metadata={"input": "User query here"}
304+
)
305+
run_id = response["run_id"]
306+
307+
# Log steps during execution
308+
client.log(run_id, "info", "Processing step 1")
309+
310+
# End the run with results
311+
client.trace(
312+
run_id=run_id,
313+
agent_id="my-agent",
314+
framework="langchain",
315+
event_type="agent_end",
316+
metadata={
317+
"input": "User query",
318+
"output": "Agent response",
319+
"steps": [
320+
{"step_number": 1, "step_name": "Analyze", "input": "...", "output": "...", "latency_ms": 1500, "status": "completed"}
321+
],
322+
"latency_ms": 1500
323+
}
324+
)</code></pre>
276325

277326
<ul class="step-list">
278327
<li>
279-
<strong>Sign In / Open Dashboard</strong>
280-
<p>Open your web browser and navigate to your Zentaxa instance URL (e.g., <code>http://localhost:5173</code>). No complex setup is required for the viewer.</p>
328+
<strong>Open the Dashboard</strong>
329+
<p>Visit <a href="https://zentaxaapp.azurewebsites.net" target="_blank" style="color: var(--primary);">zentaxaapp.azurewebsites.net</a> to access the live dashboard. No setup required.</p>
281330
</li>
282331
<li>
283-
<strong>Connect Agents</strong>
284-
<p>Your engineering team will integrate the Zentaxa SDK into your AI agents. Once connected, data will start appearing in the dashboard automatically.</p>
332+
<strong>Install SDK</strong>
333+
<p>Run <code>pip install zentaxa</code> in your Python environment. The SDK automatically connects to the live platform.</p>
285334
</li>
286335
<li>
287-
<strong>View Incoming Logs</strong>
288-
<p>Click on the <strong>"Pipeline Explorer"</strong> tab in the sidebar. You should see a list of recent activities populating in real-time.</p>
336+
<strong>Integrate with Your Agent</strong>
337+
<p>Add the ZentaxaClient to your AI agent code. Use <code>trace()</code> to track runs and <code>log()</code> to record steps.</p>
289338
</li>
290339
<li>
291-
<strong>Read Basic Metrics</strong>
292-
<p>On the main <strong>Dashboard</strong> page, look at the top cards. <strong>Success Rate</strong> should ideally be high (green). <strong>Total Cost</strong> shows your daily spend.</p>
340+
<strong>View Data in Pipeline Explorer</strong>
341+
<p>Click on <strong>"Pipeline Explorer"</strong> to see your agent runs with step-by-step visualization. Click any run to expand and see detailed input/output for each step.</p>
293342
</li>
294343
</ul>
295344
</section>
@@ -299,19 +348,28 @@ <h2>3. Getting Started</h2>
299348
<h2>4. How to Use Each Module</h2>
300349

301350
<h3>Dashboard (The Command Center)</h3>
351+
<p><strong>URL:</strong> <a href="https://zentaxaapp.azurewebsites.net/dashboard" target="_blank" style="color: var(--primary);">zentaxaapp.azurewebsites.net/dashboard</a></p>
302352
<p><strong>Purpose:</strong> High-level health check of your entire AI system.</p>
303353
<p><strong>What you see:</strong> Graphs showing cost trends, success vs. failure rates, and a list of currently active agents.</p>
304354
<p><strong>Typical Use Case:</strong> Check this page first thing in the morning to ensure no critical failures occurred overnight and that costs are within budget.</p>
305355

306-
<h3>Pipeline Explorer (The Detective Tool)</h3>
307-
<p><strong>Purpose:</strong> Deep dive into specific actions and requests.</p>
308-
<p><strong>What you see:</strong> A detailed table of every single interaction with an AI model. You can filter by date, agent name, or status (Success/Error).</p>
309-
<p><strong>Typical Use Case:</strong> A user reports a "weird answer" from the chatbot. You search here for the specific conversation to see exactly what the AI was asked and how it responded.</p>
310-
311-
<h3>Agent Runs (The Storyteller)</h3>
312-
<p><strong>Purpose:</strong> Understand full workflows and multi-step tasks.</p>
313-
<p><strong>What you see:</strong> Grouped actions that belong to a single task. For example, a "Research Task" might involve 5 separate AI calls. This view groups them together.</p>
314-
<p><strong>Typical Use Case:</strong> Debugging a complex agent that got stuck in a loop while trying to search the web.</p>
356+
<h3>Pipeline Explorer (Multi-Step Visualization)</h3>
357+
<p><strong>URL:</strong> <a href="https://zentaxaapp.azurewebsites.net/pipeline" target="_blank" style="color: var(--primary);">zentaxaapp.azurewebsites.net/pipeline</a></p>
358+
<p><strong>Purpose:</strong> Visualize multi-step agent workflows with expandable step details.</p>
359+
<p><strong>What you see:</strong> A timeline view of agent runs. Click any run to see a vertical step-by-step visualization with input prompts and LLM outputs for each step.</p>
360+
<p><strong>Typical Use Case:</strong> Debug complex multi-step agents by examining exactly what prompt was sent and what response was received at each step.</p>
361+
362+
<h3>Requests Hub (The Detective Tool)</h3>
363+
<p><strong>URL:</strong> <a href="https://zentaxaapp.azurewebsites.net/requests" target="_blank" style="color: var(--primary);">zentaxaapp.azurewebsites.net/requests</a></p>
364+
<p><strong>Purpose:</strong> Overview of all agent runs with quick access to input/output.</p>
365+
<p><strong>What you see:</strong> A list of every agent run with status, latency, and cost. Expand any run to see the full input and output.</p>
366+
<p><strong>Typical Use Case:</strong> A user reports a "weird answer" from the chatbot. Search here for the specific conversation to see exactly what the AI was asked and how it responded.</p>
367+
368+
<h3>Connections (Agent Management)</h3>
369+
<p><strong>URL:</strong> <a href="https://zentaxaapp.azurewebsites.net/connections" target="_blank" style="color: var(--primary);">zentaxaapp.azurewebsites.net/connections</a></p>
370+
<p><strong>Purpose:</strong> Manage and monitor connected agents.</p>
371+
<p><strong>What you see:</strong> List of all registered agents with their status and last activity.</p>
372+
<p><strong>Typical Use Case:</strong> Check which agents are active and verify new integrations are reporting data correctly.</p>
315373
</section>
316374

317375
<!-- 5. Troubleshooting Guide -->
@@ -350,31 +408,118 @@ <h2>6. Best Practices</h2>
350408
<h2>7. Glossary</h2>
351409
<dl style="display: grid; grid-template-columns: 150px 1fr; gap: 1rem; row-gap: 1.5rem;">
352410
<dt class="glossary-term">Trace</dt>
353-
<dd>A complete record of a single operation or request from start to finish.</dd>
411+
<dd>A complete record of a single operation or request from start to finish. Use <code>client.trace()</code> to record agent lifecycle events.</dd>
354412

355-
<dt class="glossary-term">Node</dt>
356-
<dd>A single step or component in an AI workflow (e.g., "Search Tool" or "Summarizer").</dd>
413+
<dt class="glossary-term">Run</dt>
414+
<dd>A single execution session of an agent performing a specific task. Each run has a unique <code>run_id</code>.</dd>
357415

358-
<dt class="glossary-term">Event</dt>
359-
<dd>Something noteworthy that happened, like "Tool Started" or "Error Occurred".</dd>
416+
<dt class="glossary-term">Step</dt>
417+
<dd>A single action within a multi-step agent workflow. Steps are stored in <code>meta_data.steps</code> and visualized in Pipeline Explorer.</dd>
360418

361419
<dt class="glossary-term">Agent</dt>
362-
<dd>An autonomous AI program that performs tasks and makes decisions.</dd>
420+
<dd>An autonomous AI program that performs tasks and makes decisions. Identified by <code>agent_id</code> in the SDK.</dd>
421+
422+
<dt class="glossary-term">Framework</dt>
423+
<dd>The AI framework used (e.g., LangChain, CrewAI, AutoGen, LlamaIndex, or custom).</dd>
363424

364425
<dt class="glossary-term">Latency</dt>
365-
<dd>The time delay between a request and a response (speed). Lower is better.</dd>
426+
<dd>The time delay between a request and a response in milliseconds. Stored as <code>latency_ms</code>.</dd>
366427

367428
<dt class="glossary-term">Token Cost</dt>
368-
<dd>The price paid to AI providers (like OpenAI) based on the amount of text processed.</dd>
429+
<dd>The price paid to AI providers (like OpenAI, Gemini) based on the amount of text processed.</dd>
369430

370-
<dt class="glossary-term">Run</dt>
371-
<dd>A single execution session of an agent performing a specific task.</dd>
431+
<dt class="glossary-term">SDK</dt>
432+
<dd>Software Development Kit. Install with <code>pip install zentaxa</code> to integrate observability into your agents.</dd>
372433

373-
<dt class="glossary-term">API Call</dt>
374-
<dd>A message sent to an external service to get a response or data.</dd>
434+
<dt class="glossary-term">ZentaxaClient</dt>
435+
<dd>The main SDK class for sending telemetry. Methods include <code>trace()</code>, <code>log()</code>, and <code>metric()</code>.</dd>
375436
</dl>
376437
</section>
377438

439+
<!-- 8. SDK Reference -->
440+
<section id="sdk-reference">
441+
<h2>8. SDK Reference</h2>
442+
443+
<h3>Installation</h3>
444+
<pre style="background: var(--surface); padding: 1rem; border-radius: var(--radius); border: 1px solid var(--border); overflow-x: auto;"><code># Basic installation
445+
pip install zentaxa
446+
447+
# With framework integrations
448+
pip install zentaxa[langchain]
449+
pip install zentaxa[crewai]
450+
pip install zentaxa[all]</code></pre>
451+
452+
<h3>Core Methods</h3>
453+
<div class="feature-grid">
454+
<div class="feature-card">
455+
<h4>trace()</h4>
456+
<p>Record agent lifecycle events (start, end, error). Returns <code>run_id</code> for new runs.</p>
457+
<pre style="font-size: 0.8rem; margin-top: 0.5rem;"><code>client.trace(
458+
run_id=None, # None for new run
459+
agent_id="my-agent",
460+
framework="langchain",
461+
event_type="agent_start",
462+
metadata={"input": "..."}
463+
)</code></pre>
464+
</div>
465+
<div class="feature-card">
466+
<h4>log()</h4>
467+
<p>Send log messages with level (info, warning, error).</p>
468+
<pre style="font-size: 0.8rem; margin-top: 0.5rem;"><code>client.log(
469+
run_id,
470+
"info",
471+
"Processing step 1"
472+
)</code></pre>
473+
</div>
474+
<div class="feature-card">
475+
<h4>metric()</h4>
476+
<p>Record numeric metrics like latency, cost, or tokens.</p>
477+
<pre style="font-size: 0.8rem; margin-top: 0.5rem;"><code>client.metric(
478+
run_id,
479+
"llm_latency_ms",
480+
1500.0,
481+
tags={"model": "gpt-4"}
482+
)</code></pre>
483+
</div>
484+
</div>
485+
486+
<h3>Multi-Step Agents</h3>
487+
<p>To visualize multi-step workflows in Pipeline Explorer, include a <code>steps</code> array in your final trace metadata:</p>
488+
<pre style="background: var(--surface); padding: 1rem; border-radius: var(--radius); border: 1px solid var(--border); overflow-x: auto;"><code>client.trace(
489+
run_id=run_id,
490+
agent_id="my-agent",
491+
framework="custom",
492+
event_type="agent_end",
493+
metadata={
494+
"input": "Original query",
495+
"output": "Final response",
496+
"steps": [
497+
{
498+
"step_number": 1,
499+
"step_name": "Analyze",
500+
"input": "What is machine learning?",
501+
"output": "Machine learning is...",
502+
"latency_ms": 1500,
503+
"status": "completed"
504+
},
505+
{
506+
"step_number": 2,
507+
"step_name": "Research",
508+
"input": "Find examples of ML",
509+
"output": "Examples include...",
510+
"latency_ms": 1200,
511+
"status": "completed"
512+
}
513+
],
514+
"latency_ms": 2700,
515+
"model": "gemini-2.0-flash"
516+
}
517+
)</code></pre>
518+
519+
<h3>PyPI Package</h3>
520+
<p>View the package on PyPI: <a href="https://pypi.org/project/zentaxa/" target="_blank" style="color: var(--primary);">pypi.org/project/zentaxa</a></p>
521+
</section>
522+
378523
</main>
379524
</div>
380525
</div>

0 commit comments

Comments
 (0)