pentatonic-ai

AI Agent SDK

Observability, memory, and analytics for LLM applications.
Provider-agnostic. JavaScript & Python.

---

Overview

The Pentatonic AI Agent SDK instruments your LLM applications with zero-config observability. Wrap any OpenAI, Anthropic, or Cloudflare Workers AI client and get:

• Conversation tracking -- every LLM call emits structured events (token usage, tool calls, model, latency)
• Shared memory -- semantic search across your team's AI interactions
• Session analytics -- conversation flows, dead-end detection, search-to-click metrics
• Pattern detection -- Bayesian analysis of recurring behaviors across your event stream
• Claude Code plugin -- automatic tracking for Claude Code sessions via hooks

Quick Start

1. Create an account

npx @pentatonic-ai/ai-agent-sdk init

This walks you through account creation, email verification, and API key generation. You'll get:

TES_ENDPOINT=https://your-company.api.pentatonic.com
TES_CLIENT_ID=your-company
TES_API_KEY=tes_your-company_xxxxx

2. Install

npm install @pentatonic-ai/ai-agent-sdk

pip install pentatonic-ai-agent-sdk

3. Wrap your LLM client

JavaScript

import { TESClient } from "@pentatonic-ai/ai-agent-sdk";

const tes = new TESClient({
  clientId: process.env.TES_CLIENT_ID,
  apiKey: process.env.TES_API_KEY,
  endpoint: process.env.TES_ENDPOINT,
});

// Auto-instruments every create() call
const ai = tes.wrap(new OpenAI(), { sessionId: "conv-123" });
const result = await ai.chat.completions.create({
  model: "gpt-4o",
  messages: [{ role: "user", content: "Hello!" }],
});

Python

from pentatonic_agent_events import TESClient

tes = TESClient(
    client_id=os.environ["TES_CLIENT_ID"],
    api_key=os.environ["TES_API_KEY"],
    endpoint=os.environ["TES_ENDPOINT"],
)

ai = tes.wrap(OpenAI(), session_id="conv-123")
result = ai.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello!"}],
)

That's it. Every call emits a CHAT_TURN event with token usage, tool calls, and model info.

Supported Providers

Provider	Detection	Intercepted Method
OpenAI	client.chat.completions.create	chat.completions.create()
Anthropic	client.messages.create	messages.create()
Workers AI	client.run (JS only)	run()

All other methods pass through unchanged.

Tool-Calling Loops

For multi-round tool loops, just keep calling the wrapped client. Each call emits its own event, linked by sessionId:

const ai = tes.wrap(new OpenAI(), { sessionId: "conv-101" });

// Round 1: AI requests a tool call
const r1 = await ai.chat.completions.create({
  model: "gpt-4o",
  messages: [{ role: "user", content: "Find me running shoes" }],
  tools: [searchTool],
});

// Execute tool, feed results back...

// Round 2: AI responds with final answer
const r2 = await ai.chat.completions.create({
  model: "gpt-4o",
  messages: [...messages, { role: "tool", content: toolResult }],
});

// No manual emit needed. Both events share sessionId "conv-101".

Manual Session Control

If you need full control over when events are emitted:

const session = tes.session({ sessionId: "conv-123" });

const response = await openai.chat.completions.create({
  model: "gpt-4o",
  messages: [{ role: "user", content: "What is 2+2?" }],
});

session.record(response);
await session.emitChatTurn({
  userMessage: "What is 2+2?",
  assistantResponse: response.choices[0].message.content,
});

Claude Code Plugin

Track every Claude Code conversation automatically with shared team memory.

Install via marketplace

/plugin marketplace add Pentatonic-Ltd/ai-agent-sdk