> ## Documentation Index
> Fetch the complete documentation index at: https://arize-ax.mintlify.dev/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# BeeAI

> Trace BeeAI Framework agents with OpenInference and send spans to Arize AX for LLM observability.

[BeeAI Framework](https://github.com/i-am-bee/beeai-framework) is a TypeScript / Python framework from IBM for building production-grade AI agents — tools, memory, multi-step reasoning, and pluggable LLM backends. Arize AX captures every BeeAI agent run via the [`@arizeai/openinference-instrumentation-beeai`](https://github.com/Arize-ai/openinference/tree/main/js/packages/openinference-instrumentation-beeai) package.

## Prerequisites

* Node.js 18+
* An Arize AX account ([sign up](https://arize.com/sign-up/))
* An `OPENAI_API_KEY` from the [OpenAI Platform](https://platform.openai.com/api-keys)

## Launch Arize AX

1. Sign in to your [Arize AX account](https://app.arize.com/).
2. From **Space Settings**, copy your **Space ID** and **API Key**. You will set them as `ARIZE_SPACE_ID` and `ARIZE_API_KEY` below.

## Install

```bash theme={null}
npm install "beeai-framework@0.1.13" "@ai-sdk/openai" \
  @arizeai/openinference-instrumentation-beeai \
  @arizeai/openinference-semantic-conventions \
  @opentelemetry/exporter-trace-otlp-proto \
  @opentelemetry/instrumentation \
  @opentelemetry/resources \
  @opentelemetry/sdk-trace-base \
  @opentelemetry/sdk-trace-node \
  @opentelemetry/semantic-conventions
```

<Note>
  `@arizeai/openinference-instrumentation-beeai` only patches `beeai-framework` versions in the range `>=0.1.9 <0.1.14` (declared as `INSTRUMENTS` in the instrumentor source); the install command above pins to `0.1.13` to stay inside that range. Newer `beeai-framework` releases install fine but the instrumentor's `manuallyInstrument(...)` returns silently without patching, so no spans are emitted. Pin **exactly** rather than `^` or `~` — both resolve to the latest matching release, which today is `0.1.29` (out of range). The `@ai-sdk/openai` package is required because BeeAI declares it as an *optional* peer dependency — the example uses the OpenAI provider, so it has to be installed explicitly. Track [`Arize-ai/openinference`](https://github.com/Arize-ai/openinference/tree/main/js/packages/openinference-instrumentation-beeai) for an instrumentor update that supports newer `beeai-framework` releases; once published, drop the version pin.
</Note>

## Configure credentials

```bash theme={null}
export ARIZE_SPACE_ID="<your-space-id>"
export ARIZE_API_KEY="<your-api-key>"
export ARIZE_PROJECT_NAME="beeai-tracing-example"
export OPENAI_API_KEY="<your-openai-api-key>"
```

## Setup tracing

```typescript theme={null}
// instrumentation.ts
import { OTLPTraceExporter } from "@opentelemetry/exporter-trace-otlp-proto";
import { resourceFromAttributes } from "@opentelemetry/resources";
import { SimpleSpanProcessor } from "@opentelemetry/sdk-trace-base";
import { NodeTracerProvider } from "@opentelemetry/sdk-trace-node";
import { ATTR_SERVICE_NAME } from "@opentelemetry/semantic-conventions";
import {
  SEMRESATTRS_PROJECT_NAME,
} from "@arizeai/openinference-semantic-conventions";
import { registerInstrumentations } from "@opentelemetry/instrumentation";
import {
  BeeAIInstrumentation,
} from "@arizeai/openinference-instrumentation-beeai";
import * as beeaiFramework from "beeai-framework";

const projectName =
  process.env.ARIZE_PROJECT_NAME ?? "beeai-tracing-example";

export const provider = new NodeTracerProvider({
  resource: resourceFromAttributes({
    [ATTR_SERVICE_NAME]: projectName,
    [SEMRESATTRS_PROJECT_NAME]: projectName,
  }),
  spanProcessors: [
    new SimpleSpanProcessor(
      new OTLPTraceExporter({
        url: "https://otlp.arize.com/v1/traces",
        headers: {
          "arize-space-id": process.env.ARIZE_SPACE_ID ?? "",
          "arize-api-key": process.env.ARIZE_API_KEY ?? "",
        },
      }),
    ),
  ],
});

provider.register();

const instrumentation = new BeeAIInstrumentation();
instrumentation.manuallyInstrument(beeaiFramework);

registerInstrumentations({ instrumentations: [instrumentation] });

console.log("Arize AX tracing initialized for BeeAI.");
```

## Run BeeAI

```typescript theme={null}
// example.ts

// Importing instrumentation first ensures tracing is set up before any
// BeeAI agent is created.
import { provider } from "./instrumentation";

import {
  ToolCallingAgent,
} from "beeai-framework/agents/toolCalling/agent";
import { TokenMemory } from "beeai-framework/memory/tokenMemory";
import {
  OpenAIChatModel,
} from "beeai-framework/adapters/openai/backend/chat";

// OpenAIChatModel reads OPENAI_API_KEY from the environment.
const llm = new OpenAIChatModel("gpt-5");

const agent = new ToolCallingAgent({
  llm,
  memory: new TokenMemory(),
  tools: [],
});

const response = await agent.run({
  prompt: "Why is the ocean salty? Answer in two sentences.",
});

console.log(response.result.text);

// Flush any pending spans before the process exits.
await provider.forceFlush();
```

### Expected output

```text wrap theme={null}
Arize AX tracing initialized for BeeAI.
The ocean is salty because rivers continuously dissolve mineral salts from rocks and soil and carry them to the sea, where they accumulate over millions of years. Water leaves the ocean through evaporation but the salts remain, steadily concentrating until reaching today's roughly 3.5% salinity.
```

## Verify in Arize AX

1. Open your Arize AX space and select project **`beeai-tracing-example`**.
2. You should see a new trace within \~30 seconds with this shape: a `beeai-framework-main` root span (AGENT) wraps `agent.toolCalling.start-1` / `agent.toolCalling.success-1` (AGENT), `tool.dynamic.finalAnswer.{start,finish,success}-1` (TOOL), and `backend.openai.chat.{start,finish,success}-1` (LLM, model `gpt-5`) child spans. Prompts, responses, and token usage are attached to the LLM spans.
3. If no traces appear, see [Troubleshooting](#troubleshooting).

## Troubleshooting

* **No traces in Arize AX.** Confirm `ARIZE_SPACE_ID` and `ARIZE_API_KEY` are set in the same shell that runs `example.ts`. Enable OpenTelemetry debug logs with `export OTEL_LOG_LEVEL=debug` and re-run.
* **BeeAI spans missing.** `instrumentation.manuallyInstrument(beeaiFramework)` must run before any agent is constructed. Make sure `import { provider } from "./instrumentation"` is the first import in your entry point.
* **`401` from OpenAI.** Verify `OPENAI_API_KEY` is set and has access to `gpt-5`. Swap for a model your key can call.
* **Example runs but no spans appear.** The instrumentor only patches `beeai-framework` versions in its declared `INSTRUMENTS` range (currently `>=0.1.9 <0.1.14`). Newer or older releases install fine but `manuallyInstrument(...)` silently no-ops. Pin to a supported version (the install command above pins to `~0.1.13`), or enable OpenTelemetry diag logs (`export OTEL_LOG_LEVEL=debug`) to see the `DependencyConflict` warning the instrumentor emits at startup.
* **`Cannot find package '@ai-sdk/openai'` at runtime.** BeeAI declares the `@ai-sdk/*` provider packages as *optional* peer dependencies — the user has to install only the ones they use. The example uses OpenAI; if you swap to Anthropic, install `@ai-sdk/anthropic` instead.
* **Process exits before spans flush.** Spans are exported asynchronously; always `await provider.forceFlush()` (or `provider.shutdown()`) before the process exits to avoid losing trailing spans.

## Resources

<CardGroup>
  <Card icon="book-open" href="https://framework.beeai.dev/" title="BeeAI Framework Documentation" horizontal />

  <Card icon="terminal" href="https://github.com/Arize-ai/openinference/tree/main/js/packages/openinference-instrumentation-beeai" title="OpenInference BeeAI Instrumentor (JS/TS)" horizontal />

  <Card icon="github" href="https://github.com/i-am-bee/beeai-framework" title="BeeAI Framework GitHub" horizontal />
</CardGroup>