Overmind

https://docs.overmindlab.ai
Admin
Overmind is an autonomous optimization platform for production AI agents that analyzes code,...
Tokens:106,461
Snippets:600
Trust Score:5.4
Update:5 months ago
Show doc for...
Tokens:
Raw
### Provider Examples

Source: https://docs.overmindlab.ai/guides/sdk-python

Examples of initializing the SDK with different LLM providers.

```APIDOC
from overmind import init
from openai import OpenAI


init(service_name="my-service", providers=["openai"])


client = OpenAI()
response = client.chat.completions.create(
    model="gpt-5-mini",
    messages=[{"role": "user", "content": "What is quantum computing?"}],
)
print(response.choices[0].message.content)

# Anthropic
from overmind import init
import anthropic


init(service_name="my-service", providers=["anthropic"])


client = anthropic.Anthropic()
message = client.messages.create(
    model="claude-opus-4-5",
    max_tokens=1024,
    messages=[{"role": "user", "content": "What is quantum computing?"}],
)
print(message.content[0].text)

# Google Gemini
from overmind import init
from google import genai


init(service_name="my-service", providers=["google"])


client = genai.Client()
response = client.models.generate_content(
    model="gemini-2.0-flash",
    contents="What is quantum computing?",
)
print(response.text)

# Agno
from overmind import init
from agno.agent import Agent
from agno.models.openai import OpenAIChat


init(service_name="my-service", providers=["agno"])


agent = Agent(model=OpenAIChat(id="gpt-5"), markdown=True)
agent.print_response("Write a haiku about the ocean.")

# Auto-detect all installed providers
from overmind import init


init(service_name="my-service")  # auto-instruments every supported package that is installed
```

--------------------------------

### Agno Provider Example

Source: https://docs.overmindlab.ai/guides/sdk-python

Initialize the SDK for Agno and use the Agno Agent to get a response. The SDK automatically traces this call.

```python
from overmind import init
from agno.agent import Agent
from agno.models.openai import OpenAIChat


init(service_name="my-service", providers=["agno"])


agent = Agent(model=OpenAIChat(id="gpt-5"), markdown=True)
agent.print_response("Write a haiku about the ocean.")
```

--------------------------------

### Installation

Source: https://docs.overmindlab.ai/guides/sdk-python

Install the Overmind SDK and optional provider integrations.

```APIDOC
pip install overmind

pip install overmind openai          # OpenAI
pip install overmind anthropic       # Anthropic
pip install overmind google-genai    # Google Gemini
pip install overmind agno            # Agno
```

--------------------------------

### Install SDK and Providers

Source: https://docs.overmindlab.ai/guides/sdk-js

Install the Overmind tracing SDK along with the specific LLM provider packages you intend to use.

```bash
# OpenAI
npm install @overmind-lab/trace-sdk openai


# Anthropic
npm install @overmind-lab/trace-sdk @anthropic-ai/sdk


# Google Gemini
npm install @overmind-lab/trace-sdk @google/genai
```

--------------------------------

### Install Overmind Lab with aiohttp Backend

Source: https://docs.overmindlab.ai/api/python

Install the library with the 'aiohttp' extra to enable it as an HTTP backend for improved concurrency. This command installs from the staging repository.

```bash
pip install 'overmind_lab[aiohttp] @ git+ssh://git@github.com/stainless-sdks/overmind-lab-python.git'
```

--------------------------------

### Install and Initialize Overmind

Source: https://docs.overmindlab.ai/guides/skills

Install the Overmind package using pip and initialize your agent project. This sets up the necessary configuration and installs IDE skills.

```bash
pip install overmind
overmind init
```

--------------------------------

### Install and Initialize Overmind

Source: https://docs.overmindlab.ai/guides/skills

Run this command once per project to install the Overmind package and initialize the project structure. Ensure you are in the project directory.

```bash
pip install overmind && overmind init
```

--------------------------------

### Install Overmind and OpenAI SDK (Python)

Source: https://docs.overmindlab.ai/guides/how-to-use-tracing

Install the necessary Python packages for Overmind tracing and OpenAI integration.

```bash
pip install overmind openai

```

--------------------------------

### Full Overmind SDK Example

Source: https://docs.overmindlab.ai/guides/sdk-python

This example demonstrates initializing the Overmind SDK, setting user context and tags, and capturing exceptions within a traced function. Call `init()` once at the top of your entry point. Use `service_name` meaningfully for different agents. Tag traces with context using `set_user()` and `set_tag()`.

```python
import os
from overmind import init, get_tracer, set_user, set_tag, capture_exception
from openai import OpenAI


os.environ["OVERMIND_API_KEY"] = "ovr_your_key_here"


init(
    service_name="customer-support",
    environment="production",
    providers=["openai"],
)


client = OpenAI()


def handle_support_query(user_id: str, question: str) -> str:
    set_user(user_id=user_id)
    set_tag("workflow", "support")


    tracer = get_tracer()
    with tracer.start_as_current_span("handle-query"):
        try:
            response = client.chat.completions.create(
                model="gpt-5-mini",
                messages=[
                    {"role": "system", "content": "You are a helpful customer support agent."},
                    {"role": "user", "content": question},
                ],
            )
            return response.choices[0].message.content
        except Exception as e:
            capture_exception(e)
            raise


answer = handle_support_query("user-123", "How do I reset my password?")
print(answer)
```

--------------------------------

### Install Overmind SDK

Source: https://docs.overmindlab.ai/guides/sdk-python

Install the Overmind SDK using pip. Additional packages are available for specific LLM providers.

```bash
pip install overmind
```

```bash
pip install overmind openai          # OpenAI
pip install overmind anthropic       # Anthropic
pip install overmind google-genai    # Google Gemini
pip install overmind agno            # Agno
```

--------------------------------

### Project Creation Response Example

Source: https://docs.overmindlab.ai/api/python/resources/projects/methods/create

This is an example of the JSON response received after successfully creating a project. It includes details like ID, timestamps, name, slug, and optional fields.

```json
{
  "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "created_at": "2019-12-27T18:11:19.117Z",
  "name": "name",
  "slug": "slug",
  "updated_at": "2019-12-27T18:11:19.117Z",
  "description": "description",
  "is_active": true,
  "settings": {}
}
```

--------------------------------

### Membership Add Response Example

Source: https://docs.overmindlab.ai/api/python/resources/projects/subresources/memberships/methods/add

This is an example of the JSON response received after successfully adding a user to a project membership.

```json
{
  "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "created_at": "2019-12-27T18:11:19.117Z",
  "project": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "updated_at": "2019-12-27T18:11:19.117Z",
  "user": 0,
  "user_email": "dev@stainless.com"
}
```

--------------------------------

### Usage Example

Source: https://docs.overmindlab.ai/api/typescript

Demonstrates how to initialize the client and make a basic authenticated request to retrieve the current user.

```APIDOC
## Usage Example

This snippet shows how to instantiate the OvermindLab client and call the `retrieveCurrentUser` method.

### Method
```typescript
client.auth.retrieveCurrentUser()
```

### Request Example
```typescript
import OvermindLab from 'overmind-lab';

const client = new OvermindLab({
  environment: 'environment_1', // defaults to 'production'
});

const response = await client.auth.retrieveCurrentUser();

console.log(response.id);
```

### Response Example
```json
{
  "id": "user_id_123",
  "name": "John Doe"
}
```
```

--------------------------------

### Token Creation Response Example

Source: https://docs.overmindlab.ai/api/python/resources/auth/subresources/token/methods/create

This is an example of a successful response when creating authentication tokens, containing the access and refresh tokens.

```json
{
  "access": "access",
  "refresh": "refresh"
}
```

--------------------------------

### Job Iteration Creation Response Example

Source: https://docs.overmindlab.ai/api/python/resources/job_iterations/methods/create

This is an example of the JSON response received after successfully creating a job iteration.

```json
{
  "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "created_at": "2019-12-27T18:11:19.117Z",
  "iteration_name": "iteration_name",
  "job": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "agent_code": "agent_code",
  "avg_score": 0,
  "description": "description",
  "dimension_scores": {},
  "order": -9007199254740991,
  "status": "keep"
}
```

--------------------------------

### Register Response Example

Source: https://docs.overmindlab.ai/api/python/resources/auth/methods/register

This is an example of a successful response when registering a new user. It includes access and refresh tokens, along with user details.

```json
{
  "access": "access",
  "refresh": "refresh",
  "user": {
    "id": 0,
    "email": "dev@stainless.com",
    "email_verified": true
  }
}
```

--------------------------------

### Create Job Response Example

Source: https://docs.overmindlab.ai/api/resources/jobs/methods/create

Example of a successful response when creating a job.

```APIDOC
## Response Example

### Success Response (200)

- **id** (string) - Unique identifier for the job.
- **created_at** (string) - Timestamp when the job was created.
- **iterations** (array) - List of iterations for the job.
  - **id** (string) - Unique identifier for the iteration.
  - **created_at** (string) - Timestamp when the iteration was created.
  - **iteration_name** (string) - Name of the iteration.
  - **job** (string) - Identifier of the associated job.
  - **agent_code** (string) - Code of the agent used in the iteration.
  - **avg_score** (number) - Average score of the iteration.
  - **description** (string) - Description of the iteration.
  - **dimension_scores** (object) - Scores for different dimensions.
  - **order** (number) - Order of the iteration.
  - **status** (string) - Status of the iteration (e.g., "keep").
- **updated_at** (string) - Timestamp when the job was last updated.
- **agent** (string) - Identifier of the agent used for the job.
- **analyzer_model** (string) - The analyzer model used.
- **backtest_results** (object) - Results from backtesting.
- **baseline_score** (number) - The baseline score for the job.
- **best_agent_code** (string) - The agent code that achieved the best score.
- **best_score** (number) - The best score achieved.
- **candidates_per_iteration** (number) - Number of candidates per iteration.
- **celery_task_id** (string) - Celery task ID associated with the job.
- **data_source** (string) - The data source used for the job.
- **improvement** (number) - Improvement achieved by the job.
- **job_type** (string) - Type of the job (e.g., "inference").
- **num_iterations** (number) - Total number of iterations.
- **project** (string) - Identifier of the project the job belongs to.
- **prompt_slug** (string) - Slug for the prompt used.
- **report_markdown** (string) - Markdown report of the job results.
- **result** (object) - The result of the job.
- **status** (string) - Current status of the job (e.g., "pending").
- **triggered_by** (number) - Identifier of who or what triggered the job.

### Response Example

```json
{
  "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "created_at": "2019-12-27T18:11:19.117Z",
  "iterations": [
    {
      "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "created_at": "2019-12-27T18:11:19.117Z",
      "iteration_name": "iteration_name",
      "job": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "agent_code": "agent_code",
      "avg_score": 0,
      "description": "description",
      "dimension_scores": {},
      "order": -9007199254740991,
      "status": "keep"
    }
  ],
  "updated_at": "2019-12-27T18:11:19.117Z",
  "agent": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "analyzer_model": "analyzer_model",
  "backtest_results": {},
  "baseline_score": 0,
  "best_agent_code": "best_agent_code",
  "best_score": 0,
  "candidates_per_iteration": -9007199254740991,
  "celery_task_id": "celery_task_id",
  "data_source": "data_source",
  "improvement": 0,
  "job_type": "inference",
  "num_iterations": -9007199254740991,
  "project": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "prompt_slug": "prompt_slug",
  "report_markdown": "report_markdown",
  "result": {},
  "status": "pending",
  "triggered_by": 0
}
```
```

--------------------------------

### Job Creation Response Example

Source: https://docs.overmindlab.ai/api/typescript/resources/jobs/methods/create

Example of a successful response when creating a job.

```APIDOC
## POST /resources/jobs

### Description
Creates a new job.

### Method
POST

### Endpoint
/resources/jobs

### Response
#### Success Response (200)
- **id** (string) - Unique identifier for the job.
- **created_at** (string) - Timestamp when the job was created.
- **iterations** (array) - List of iterations for the job.
  - **id** (string) - Unique identifier for the iteration.
  - **created_at** (string) - Timestamp when the iteration was created.
  - **iteration_name** (string) - Name of the iteration.
  - **job** (string) - ID of the job this iteration belongs to.
  - **agent_code** (string) - Code of the agent used in this iteration.
  - **avg_score** (number) - Average score for the iteration.
  - **description** (string) - Description of the iteration.
  - **dimension_scores** (object) - Scores for different dimensions.
  - **order** (number) - Order of the iteration.
  - **status** (string) - Status of the iteration (e.g., 'keep').
- **updated_at** (string) - Timestamp when the job was last updated.
- **agent** (string) - Identifier for the agent used.
- **analyzer_model** (string) - The analyzer model used.
- **backtest_results** (object) - Results from backtesting.
- **baseline_score** (number) - The baseline score.
- **best_agent_code** (string) - The code for the best agent.
- **best_score** (number) - The best score achieved.
- **candidates_per_iteration** (number) - Number of candidates per iteration.
- **celery_task_id** (string) - Celery task ID associated with the job.
- **data_source** (string) - The data source used for the job.
- **improvement** (number) - The improvement achieved.
- **job_type** (string) - The type of job (e.g., 'inference').
- **num_iterations** (number) - The total number of iterations.
- **project** (string) - Identifier for the project.
- **prompt_slug** (string) - Slug for the prompt.
- **report_markdown** (string) - Markdown report of the job results.
- **result** (object) - The result of the job.
- **status** (string) - The current status of the job (e.g., 'pending').
- **triggered_by** (number) - Identifier for who triggered the job.

#### Response Example
```json
{
  "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "created_at": "2019-12-27T18:11:19.117Z",
  "iterations": [
    {
      "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "created_at": "2019-12-27T18:11:19.117Z",
      "iteration_name": "iteration_name",
      "job": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "agent_code": "agent_code",
      "avg_score": 0,
      "description": "description",
      "dimension_scores": {},
      "order": -9007199254740991,
      "status": "keep"
    }
  ],
  "updated_at": "2019-12-27T18:11:19.117Z",
  "agent": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "analyzer_model": "analyzer_model",
  "backtest_results": {},
  "baseline_score": 0,
  "best_agent_code": "best_agent_code",
  "best_score": 0,
  "candidates_per_iteration": -9007199254740991,
  "celery_task_id": "celery_task_id",
  "data_source": "data_source",
  "improvement": 0,
  "job_type": "inference",
  "num_iterations": -9007199254740991,
  "project": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "prompt_slug": "prompt_slug",
  "report_markdown": "report_markdown",
  "result": {},
  "status": "pending",
  "triggered_by": 0
}
```
```

--------------------------------

### Agent Creation Example

Source: https://docs.overmindlab.ai/api/typescript

Illustrates how to create a new agent by providing the necessary parameters and handling the response.

```APIDOC
## Agent Creation Example

This example demonstrates how to create an agent using the `agents.create` method, including defining the request parameters and handling the response type.

### Method
```typescript
client.agents.create(params: OvermindLab.AgentCreateParams)
```

### Parameters
#### Request Body
- **name** (string) - Required - The name of the agent.
- **project** (string) - Required - The ID of the project the agent belongs to.
- **slug** (string) - Required - A unique slug for the agent.

### Request Example
```typescript
import OvermindLab from 'overmind-lab';

const client = new OvermindLab({
  environment: 'environment_1', // defaults to 'production'
});

const params: OvermindLab.AgentCreateParams = {
  name: 'x',
  project: '182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e',
  slug: 'slug',
};
const agent: OvermindLab.AgentCreateResponse = await client.agents.create(params);
```

### Response Example
```json
{
  "id": "agent_id_456",
  "name": "x",
  "project": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "slug": "slug"
}
```
```

--------------------------------

### Install Overmind

Source: https://docs.overmindlab.ai/guides/getting-started

Install the Overmind package using pip. Ensure you have Python 3.10 or higher.

```bash
pip install overmind
```

--------------------------------

### API Response Example (JSON)

Source: https://docs.overmindlab.ai/api/python/resources/auth/methods/retrieve_current_user

This is an example of the JSON response received when successfully retrieving the current user's profile.

```json
{
  "id": 0,
  "email": "dev@stainless.com",
  "email_verified": true
}
```

--------------------------------

### API Key Creation Response Example

Source: https://docs.overmindlab.ai/api/python/resources/auth/subresources/api_keys/methods/create

This is an example of the JSON response received after successfully creating an API key. The 'key' field is shown only once.

```json
{
  "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "created_at": "2019-12-27T18:11:19.117Z",
  "key": "key",
  "name": "name",
  "prefix": "prefix"
}
```

--------------------------------

### List Services Response Example

Source: https://docs.overmindlab.ai/api/python/resources/traces/methods/list_services

This is an example of a successful response when listing services. It includes the total count of services and a list of service names.

```json
{
  "count": 123,
  "results": [
    "string"
  ],
  "next": "http://api.example.org/accounts/?page=4",
  "previous": "http://api.example.org/accounts/?page=2"
}
```

--------------------------------

### Dataset Activate Response Example

Source: https://docs.overmindlab.ai/api/python/resources/datasets/methods/activate

This is an example of a successful JSON response when activating a dataset. It includes details about the dataset, agent, and creation timestamp.

```json
{
  "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "agent": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
  "created_at": "2019-12-27T18:11:19.117Z",
  "datapoints": [
    {
      "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "expected_output": {},
      "input": {},
      "order": 0,
      "persona": "persona",
      "tags": {}
    }
  ],
  "num_datapoints": 0,
  "version": 0,
  "generator_model": "generator_model",
  "metadata": {},
  "name": "name",
  "policy_hash": "policy_hash",
  "source": "seed"
}
```

--------------------------------

### API Response Example

Source: https://docs.overmindlab.ai/api/python/resources/projects/methods/list

This is an example of a successful response when listing projects, showing the structure of the returned data including count, results, and pagination links.

```json
{
  "count": 123,
  "results": [
    {
      "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "created_at": "2019-12-27T18:11:19.117Z",
      "name": "name",
      "slug": "slug",
      "updated_at": "2019-12-27T18:11:19.117Z",
      "description": "description",
      "is_active": true,
      "settings": {}
    }
  ],
  "next": "http://api.example.org/accounts/?page=4",
  "previous": "http://api.example.org/accounts/?page=2"
}
```

--------------------------------

### Install Overmind Trace SDK and OpenAI (JavaScript/TypeScript)

Source: https://docs.overmindlab.ai/guides/how-to-use-tracing

Install the Overmind trace SDK and the OpenAI client library using npm for your JavaScript or TypeScript project.

```bash
npm install @overmind-lab/trace-sdk openai

```

--------------------------------

### Install Overmind Lab TypeScript Library

Source: https://docs.overmindlab.ai/api

Use this command to install the Overmind Lab TypeScript client library via npm.

```bash
npm install git+ssh://git@github.com:stainless-sdks/overmind-lab-typescript.git
```

--------------------------------

### Anthropic Provider Example

Source: https://docs.overmindlab.ai/guides/sdk-python

Initialize the SDK for Anthropic and use the Anthropic client to create a message. The SDK automatically traces this call.

```python
from overmind import init
import anthropic


init(service_name="my-service", providers=["anthropic"])


client = anthropic.Anthropic()
message = client.messages.create(
    model="claude-opus-4-5",
    max_tokens=1024,
    messages=[{"role": "user", "content": "What is quantum computing?"}],
)
print(message.content[0].text)
```

--------------------------------

### Install Overmind Lab Python Library

Source: https://docs.overmindlab.ai/api/python

Install the library from the staging repository using pip. This command will be updated once the package is published to PyPI.

```bash
pip install git+ssh://git@github.com/stainless-sdks/overmind-lab-python.git
```

--------------------------------

### List Project Memberships Response Example

Source: https://docs.overmindlab.ai/api/python/resources/projects/subresources/memberships/methods/list

This is an example of the JSON response structure when successfully listing project memberships. It includes count, results, and pagination details.

```json
{
  "count": 123,
  "results": [
    {
      "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "created_at": "2019-12-27T18:11:19.117Z",
      "project": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "updated_at": "2019-12-27T18:11:19.117Z",
      "user": 0,
      "user_email": "dev@stainless.com"
    }
  ],
  "next": "http://api.example.org/accounts/?page=4",
  "previous": "http://api.example.org/accounts/?page=2"
}
```

--------------------------------

### Initialize Overmind Configuration

Source: https://docs.overmindlab.ai/

Initialize Overmind in your agent project directory. This configures API keys and installs necessary IDE skills.

```bash
cd your-agent-project/
overmind init

```

--------------------------------

### Auto-detect Providers

Source: https://docs.overmindlab.ai/guides/sdk-python

Initialize the SDK without specifying providers to automatically instrument all supported packages that are installed.

```python
from overmind import init


init(service_name="my-service")  # auto-instruments every supported package that is installed
```

--------------------------------

### Agent Policy Example

Source: https://docs.overmindlab.ai/guides/overmind_optimizer

Defines the rules and expectations for an agent's behavior, guiding the optimization process to ensure compliance with business logic.

```yaml
# Agent Policy: Lead Qualification


## Purpose
Qualifies inbound sales leads by analyzing company data and inquiry content.

```

--------------------------------

### Quick Start with OpenAI

Source: https://docs.overmindlab.ai/guides/sdk-js

Initialize the Overmind client, enable OpenAI tracing, and then use the OpenAI client as usual. Ensure your OVERMIND_API_KEY and OPENAI_API_KEY environment variables are set.

```javascript
import { OpenAI } from "openai";
import { OvermindClient } from "@overmind-lab/trace-sdk";


const overmindClient = new OvermindClient({
  apiKey: process.env.OVERMIND_API_KEY!,
  appName: "my app",
});


overmindClient.initTracing({
  enableBatching: false,
  enabledProviders: { openai: OpenAI },
});


const openai = new OpenAI({ apiKey: process.env.OPENAI_API_KEY });


const response = await openai.chat.completions.create({
  model: "gpt-5-mini",
  messages: [{ role: "user", content: "Explain quantum computing" }],
});


await overmindClient.shutdown();
```

--------------------------------

### Refresh Token Example (TypeScript)

Source: https://docs.overmindlab.ai/api/typescript/resources/auth/subresources/token/methods/refresh

Use this snippet to refresh an access token by providing a valid refresh token. Ensure you have the 'overmind-lab' SDK installed and initialized.

```typescript
import OvermindLab from 'overmind-lab';

const client = new OvermindLab();

const response = await client.auth.token.refresh({ refresh: 'x' });

console.log(response.access);
```

--------------------------------

### List Services (HTTP)

Source: https://docs.overmindlab.ai/api/resources/traces/methods/list_services

Use this endpoint to retrieve a list of all distinct service names. No specific setup is required beyond making an HTTP GET request.

```curl
curl https://api.overmindlab.ai/api/traces/services/
```

--------------------------------

### Determine Installed Version

Source: https://docs.overmindlab.ai/api/python

Use this snippet to check the currently installed version of the overmind_lab package in your Python environment. Ensure the package is installed.

```python
import overmind_lab
print(overmind_lab.__version__)
```

--------------------------------

### Initialize Overmind Project

Source: https://docs.overmindlab.ai/guides/getting-started

Navigate to your agent's project root and run 'overmind init'. This command sets up the .overmind/ directory, collects API keys, configures the analyzer model, and installs Agent Skills into your IDE.

```bash
cd your-agent-project/
overmind init
```

--------------------------------

### Quick Start with Anthropic

Source: https://docs.overmindlab.ai/guides/sdk-js

Initialize the Overmind client, enable Anthropic tracing, and then use the Anthropic client. Ensure your OVERMIND_API_KEY and ANTHROPIC_API_KEY environment variables are set.

```javascript
import * as Anthropic from "@anthropic-ai/sdk";
import { OvermindClient } from "@overmind-lab/trace-sdk";


const overmindClient = new OvermindClient({
  apiKey: process.env.OVERMIND_API_KEY!,
  appName: "my app",
});


overmindClient.initTracing({
  enableBatching: false,
  enabledProviders: { anthropic: Anthropic },
});


const client = new Anthropic.default({ apiKey: process.env.ANTHROPIC_API_KEY });


const message = await client.messages.create({
  model: "claude-sonnet-4-20250514",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Explain quantum computing" }],
});


console.log(message.content[0].text);


await overmindClient.shutdown();
```

--------------------------------

### Quick Start with Google Gemini

Source: https://docs.overmindlab.ai/guides/sdk-js

Initialize the Overmind client, enable Google Gemini tracing, and then use the Gemini client. Ensure your OVERMIND_API_KEY and GEMINI_API_KEY environment variables are set.

```javascript
import * as GoogleGenAI from "@google/genai";
import { OvermindClient } from "@overmind-lab/trace-sdk";


const overmindClient = new OvermindClient({
  apiKey: process.env.OVERMIND_API_KEY!,
  appName: "my app",
});


overmindClient.initTracing({
  enableBatching: false,
  enabledProviders: { googleGenAI: GoogleGenAI },
});


const client = new GoogleGenAI.GoogleGenAI({
  apiKey: process.env.GEMINI_API_KEY,
});


const response = await client.models.generateContent({
  model: "gemini-2.0-flash",
  contents: "Explain quantum computing",
});


console.log(response.text);


await overmindClient.shutdown();
```

--------------------------------

### Retrieve Trace Example

Source: https://docs.overmindlab.ai/api/python/resources/traces/methods/retrieve

This example shows a successful response when retrieving trace data.

```APIDOC
## GET /v1/traces

### Description
Retrieves trace data.

### Method
GET

### Endpoint
/v1/traces

### Response
#### Success Response (200)
- **root** (object) - The root span of the trace.
- **span_count** (integer) - The total number of spans in the trace.
- **spans** (array) - A list of all spans within the trace.
- **trace_id** (string) - The unique identifier for the trace.

#### Response Example
```json
{
  "root": {
    "agent": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
    "attributes": {},
    "duration_ns": 0,
    "end_time_ns": 0,
    "events": {},
    "is_root": true,
    "iteration": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
    "job": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
    "kind": 0,
    "links": {},
    "name": "name",
    "operation": "operation",
    "parent_span_id": "parent_span_id",
    "project": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
    "received_at": "2019-12-27T18:11:19.117Z",
    "resource_attrs": {},
    "scope_name": "scope_name",
    "scope_version": "scope_version",
    "service_name": "service_name",
    "span_id": "span_id",
    "span_type": "llm_call",
    "start_time_ns": 0,
    "status_code": 0,
    "status_message": "status_message",
    "trace_id": "trace_id"
  },
  "span_count": 0,
  "spans": [
    {
      "agent": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "attributes": {},
      "duration_ns": 0,
      "end_time_ns": 0,
      "events": {},
      "is_root": true,
      "iteration": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "job": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "kind": 0,
      "links": {},
      "name": "name",
      "operation": "operation",
      "parent_span_id": "parent_span_id",
      "project": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "received_at": "2019-12-27T18:11:19.117Z",
      "resource_attrs": {},
      "scope_name": "scope_name",
      "scope_version": "scope_version",
      "service_name": "service_name",
      "span_id": "span_id",
      "span_type": "llm_call",
      "start_time_ns": 0,
      "status_code": 0,
      "status_message": "status_message",
      "trace_id": "trace_id"
    }
  ],
  "trace_id": "trace_id"
}
```
```

--------------------------------

### Initialize Overmind Tracing (Python)

Source: https://docs.overmindlab.ai/guides/getting-started

Install the Overmind Python tracing SDK and call 'overmind.init()' at your application's startup to enable Overmind traces. This is useful for monitoring deployed applications independently of the optimization workflow.

```bash
pip install overmind
```

```python
import overmind
overmind.init(service_name="my-service", environment="production")
```

--------------------------------

### Google Gemini Provider Example

Source: https://docs.overmindlab.ai/guides/sdk-python

Initialize the SDK for Google Gemini and use the genai client to generate content. The SDK automatically traces this call.

```python
from overmind import init
from google import genai


init(service_name="my-service", providers=["google"])


client = genai.Client()
response = client.models.generate_content(
    model="gemini-2.0-flash",
    contents="What is quantum computing?",
)
print(response.text)
```

--------------------------------

### OpenAI Provider Example

Source: https://docs.overmindlab.ai/guides/sdk-python

Initialize the SDK for OpenAI and use the OpenAI client to make a chat completion request. The SDK automatically traces this call.

```python
from overmind import init
from openai import OpenAI


init(service_name="my-service", providers=["openai"])


client = OpenAI()
response = client.chat.completions.create(
    model="gpt-5-mini",
    messages=[{"role": "user", "content": "What is quantum computing?"}],
)
print(response.choices[0].message.content)
```

--------------------------------

### Create Agent with Basic Parameters

Source: https://docs.overmindlab.ai/api/typescript/resources/agents/methods/create

Demonstrates how to create a new agent by providing the required name, project ID, and slug. The agent's ID is logged upon successful creation.

```typescript
import OvermindLab from 'overmind-lab';

const client = new OvermindLab();

const agent = await client.agents.create({
  name: 'x',
  project: '182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e',
  slug: 'slug',
});

console.log(agent.id);
```

--------------------------------

### Agent List Response Example

Source: https://docs.overmindlab.ai/api/python/resources/agents/methods/list

Example JSON structure for a successful response when listing agents, including count and results.

```json
{
  "count": 123,
  "results": [
    {
      "id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "created_at": "2019-12-27T18:11:19.117Z",
      "name": "name",
      "project": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "slug": "slug",
      "updated_at": "2019-12-27T18:11:19.117Z",
      "agent_path": "agent_path",
      "analyzer_model": "analyzer_model",
      "dataset_has_expected_output": true,
      "dataset_size": -9007199254740991,
      "description": "description",
      "display_name": "display_name",
      "entrypoint_fn": "entrypoint_fn",
      "model": "model",
      "status": "status",
      "structure_weight": 0,
      "tags": {},
      "tool_usage_weight": 0,
      "total_points": 0
    }
  ],
  "next": "http://api.example.org/accounts/?page=4",
  "previous": "http://api.example.org/accounts/?page=2"
}
```

--------------------------------

### Initialize Overmind for Auto-Detection of Python Providers

Source: https://docs.overmindlab.ai/guides/integrations

Call `init()` once at startup. If the `providers` list is empty or omitted, all installed providers will be auto-detected and instrumented.

```python
from overmind import init


init(service_name="my-service")  # instruments all installed providers
```

--------------------------------

### Initialize Overmind with Specific Python Providers

Source: https://docs.overmindlab.ai/guides/integrations

Call `overmind.init()` once at startup, specifying the providers you intend to use. Your existing LLM client code remains unchanged.

```python
from overmind import init


init(service_name="my-service", providers=["openai", "anthropic"])
```

--------------------------------

### Agent Eval Spec Response Example

Source: https://docs.overmindlab.ai/api/python/resources/agents/methods/retrieve_eval_spec

This is an example of the JSON response structure when successfully retrieving an agent's evaluation specification.

```json
{
  "agent_description": "agent_description",
  "agent_path": "agent_path",
  "consistency_rules": {},
  "fixed_elements": {},
  "input_schema": {},
  "optimizable_elements": {},
  "output_fields": {},
  "structure_weight": 0,
  "tool_config": {},
  "tool_usage_weight": 0,
  "total_points": 0
}
```

--------------------------------

### init() Parameters

Source: https://docs.overmindlab.ai/guides/sdk-python

Parameters for the `init()` function.

```APIDOC
Parameter| Type| Default| Description  
---|---|---|---
`overmind_api_key`| `str | None`| `None`| Your Overmind API key. Falls back to `OVERMIND_API_KEY` env var.  
`service_name`| `str | None`| `None`| Name of your service, shown in the dashboard. Also reads `OVERMIND_SERVICE_NAME` env var. Defaults to `"overmind-telemetry"`.  
`environment`| `str | None`| `None`| Deployment environment (`"production"`, `"staging"`, etc.). Also reads `OVERMIND_ENVIRONMENT` env var. Defaults to `"development"`.  
`providers`| `list[str] | None`| `None`| Which LLM providers to instrument. Supported: `"openai"`, `"anthropic"`, `"google"`, `"agno"`. `None` or empty list = auto-detect all installed providers.  
`overmind_base_url`| `str | None`| `None`| Override the Overmind API endpoint. Falls back to `OVERMIND_API_URL` env var, then `https://api.overmindlab.ai`.
```

--------------------------------

### Agent Retrieval Response Example

Source: https://docs.overmindlab.ai/api/resources/agents/methods/retrieve

This is an example of a successful (200 OK) response when retrieving agent details. It includes all available fields for an agent.

```json
{
  "id": "182bd5e5-6e1a-4fe4-a799-aa6d9aab26e",
  "created_at": "2019-12-27T18:11:19.117Z",
  "name": "name",
  "project": "182bd5e5-6e1a-4fe4-a799-aa6d9aab26e",
  "slug": "slug",
  "updated_at": "2019-12-27T18:11:19.117Z",
  "active_dataset": "182bd5e5-6e1a-4fe4-a799-aa6d9aab26e",
  "agent_description": {},
  "agent_path": "agent_path",
  "analyzer_model": "analyzer_model",
  "backtest_metadata": {},
  "backtest_model_suggestions": {},
  "consistency_rules": {},
  "dataset_has_expected_output": true,
  "dataset_input_keys": {},
  "dataset_size": -9007199254740991,
  "description": "description",
  "display_name": "display_name",
  "entrypoint_fn": "entrypoint_fn",
  "eval_dataset": {},
  "evaluation_criteria": {},
  "fixed_elements": {},
  "improvement_metadata": {},
  "input_schema": {},
  "is_deleted": true,
  "model": "model",
  "optimizable_elements": {},
  "output_fields": {},
  "output_schema": {},
  "policy_data": {},
  "policy_markdown": "policy_markdown",
  "proposed_criteria": {},
  "status": "status",
  "structure_weight": 0,
  "tags": {},
  "tool_analysis": {},
  "tool_config": {},
  "tool_usage_weight": 0,
  "total_points": 0
}
```

--------------------------------

### Using Multiple Providers

Source: https://docs.overmindlab.ai/guides/sdk-js

Initialize the Overmind client and enable tracing for multiple LLM providers in a single `initTracing()` call. Batching is enabled for production use.

```javascript
import { OpenAI } from "openai";
import * as Anthropic from "@anthropic-ai/sdk";
import * as GoogleGenAI from "@google/genai";
import { OvermindClient } from "@overmind-lab/trace-sdk";


const overmindClient = new OvermindClient({
  apiKey: process.env.OVERMIND_API_KEY!,
  appName: "my app",
});


overmindClient.initTracing({
  enableBatching: true,
  enabledProviders: {
    openai: OpenAI,
    anthropic: Anthropic,
    googleGenAI: GoogleGenAI,
  },
});
```

--------------------------------

### List Projects (HTTP)

Source: https://docs.overmindlab.ai/api/resources/projects/methods/list

Use this command to list all projects. You can append query parameters to filter or order the results.

```bash
curl https://api.overmindlab.ai/api/projects/
```

--------------------------------

### Example Response for Trace Retrieval

Source: https://docs.overmindlab.ai/api/python/resources/traces/methods/retrieve

This is an example of the JSON payload returned when a trace is successfully retrieved. It includes the root span and a list of all associated spans.

```json
{
  "root": {
    "agent": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
    "attributes": {},
    "duration_ns": 0,
    "end_time_ns": 0,
    "events": {},
    "is_root": true,
    "iteration": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
    "job": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
    "kind": 0,
    "links": {},
    "name": "name",
    "operation": "operation",
    "parent_span_id": "parent_span_id",
    "project": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
    "received_at": "2019-12-27T18:11:19.117Z",
    "resource_attrs": {},
    "scope_name": "scope_name",
    "scope_version": "scope_version",
    "service_name": "service_name",
    "span_id": "span_id",
    "span_type": "llm_call",
    "start_time_ns": 0,
    "status_code": 0,
    "status_message": "status_message",
    "trace_id": "trace_id"
  },
  "span_count": 0,
  "spans": [
    {
      "agent": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "attributes": {},
      "duration_ns": 0,
      "end_time_ns": 0,
      "events": {},
      "is_root": true,
      "iteration": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "job": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "kind": 0,
      "links": {},
      "name": "name",
      "operation": "operation",
      "parent_span_id": "parent_span_id",
      "project": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
      "received_at": "2019-12-27T18:11:19.117Z",
      "resource_attrs": {},
      "scope_name": "scope_name",
      "scope_version": "scope_version",
      "service_name": "service_name",
      "span_id": "span_id",
      "span_type": "llm_call",
      "start_time_ns": 0,
      "status_code": 0,
      "status_message": "status_message",
      "trace_id": "trace_id"
    }
  ],
  "trace_id": "trace_id"
}
```

--------------------------------

### Initialize Overmind SDK

Source: https://docs.overmindlab.ai/guides/sdk-python

Initialize the SDK once at application startup. This automatically instruments existing LLM client code without requiring import changes or proxies.

```python
from overmind import init


init(
    overmind_api_key="ovr_...",
    service_name="my-service",
    environment="production",
    providers=["openai"],
)
```
Tokens:
Raw