> ## Documentation Index
> Fetch the complete documentation index at: https://docs.anchorbrowser.io/llms.txt
> Use this file to discover all available pages before exploring further.
# Authentication (OmniConnect)
> Equip your agent with a persistent identity, enabling seamless access to applications without repeated logins
## Overview
Keep **authenticated browser sessions** running for your AI agents and automations.
**Store credentials once** on an identity, and **Anchor re-authenticates automatically** when a session needs it. **Launch a browser with that identity** and your agent **starts already logged in** ready to work on the target site.
```mermaid theme={null}
flowchart LR
A["Application:
where to log in"] --> B["Auth flow:
Standard guided or custom"]
B --> C["Identity:
who logs in"]
C --> D["Browser session:
starts signed in"]
```
Define an **Application** and an **Identity**; attach that identity when you create the session so the browser starts signed in. New applications include a default **Standard authentication** flow (guided browser sign-in with automatic detection of the site's login). Add or customize **Auth Flows** when you need structured credentials or multiple login methods. For end-user sign-in inside your own product, use **[Embedding End-User Authentication UI](/essentials/omniconnect)**.
## Quick Start
```javascript node.js theme={null}
import Anchorbrowser from 'anchorbrowser';
const anchorClient = new Anchorbrowser();
// Create an application for a target website
const app = await anchorClient.applications.create({
name: 'My LinkedIn User',
source: 'linkedin.com'
});
console.log(app.id);
```
```python python theme={null}
from anchorbrowser import Anchorbrowser
anchor_client = Anchorbrowser()
# Create an application for a target website
app = anchor_client.applications.create(
name="My LinkedIn User",
source="linkedin.com"
)
print(app.id)
```
```bash cURL theme={null}
curl -X POST "https://api.anchorbrowser.io/v1/applications" \
-H "anchor-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"name": "My LinkedIn User", "source": "linkedin.com"}'
```
Applications you create already include a default **Standard authentication** flow, so you can create identities with the guided browser experience without this step. Use the API here when you need **additional** flows or structured methods—for example `username_password`, `authenticator`, `custom`, or profile-based sign-in.
```javascript node.js theme={null}
await anchorClient.applications.authFlows.create(app.id, {
name: 'Email Login with authenticator',
methods: ['username_password', 'authenticator']
});
```
```python python theme={null}
anchor_client.applications.auth_flows.create(
application_id=app.id,
name="Email Login with authenticator",
methods=["username_password", "authenticator"]
)
```
```bash cURL theme={null}
curl -X POST "https://api.anchorbrowser.io/v1/applications/{app_id}/auth-flows" \
-H "anchor-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"name": "Email Login with authenticator", "methods": ["username_password", "authenticator"]}'
```
Create an identity with credentials for the application.
```javascript node.js theme={null}
const identity = await anchorClient.identities.create({
source: 'https://linkedin.com',
name: 'John Doe',
credentials: [{
type: 'username_password',
username: 'john@example.com',
password: 'secret'
}]
});
console.log(identity.id);
```
```python python theme={null}
identity = anchor_client.identities.create(
source="https://linkedin.com",
name="John Doe",
credentials=[{
"type": "username_password",
"username": "john@example.com",
"password": "secret"
}]
)
print(identity.id)
```
```bash cURL theme={null}
curl -X POST "https://api.anchorbrowser.io/v1/identities" \
-H "anchor-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"source": "https://linkedin.com",
"name": "John Doe",
"credentials": [{
"type": "username_password",
"username": "john@example.com",
"password": "secret"
}]
}'
```
For end-user self-service authentication in your app, use [Embedding End-User Authentication UI](/essentials/omniconnect).
Use the identity ID to create an authenticated browser session.
```javascript node.js theme={null}
import Anchorbrowser from 'anchorbrowser';
const anchorClient = new Anchorbrowser();
const identityId = "your-identity-id";
const session = await anchorClient.sessions.create({
// Recommended settings for authenticated sessions.
session: {
proxy: {
active: true,
}
},
browser: {
captcha_solver: {
active: true,
},
extra_stealth: {
active: true,
}
},
// Identity to authenticate with.
identities: [{ id: identityId }]
});
console.log(session.data.id);
```
```python python theme={null}
from anchorbrowser import Anchorbrowser
anchor_client = Anchorbrowser()
identity_id = "your-identity-id"
session = anchor_client.sessions.create(
# Recommended settings for authenticated sessions.
session={
"proxy": {
"active": True,
}
},
browser={
"captcha_solver": {
"active": True,
},
"extra_stealth": {
"active": True,
}
},
# Identity to authenticate with.
identities=[{"id": identity_id}]
)
print(session.data.id)
```
## Identity Metadata
You can add custom metadata to identities for filtering and organization. Metadata is a flexible key-value store that allows you to tag identities with custom attributes.
### Creating an Identity with Metadata
```javascript node.js theme={null}
const identity = await anchorClient.identities.create({
source: 'https://linkedin.com',
name: 'John Doe',
metadata: {
department: 'Engineering',
role: 'admin'
},
credentials: [{
type: 'username_password',
username: 'john@example.com',
password: 'secret'
}]
});
```
```python python theme={null}
identity = anchor_client.identities.create(
source="https://linkedin.com",
name="John Doe",
metadata={
"department": "Engineering",
"role": "admin"
},
credentials=[{
"type": "username_password",
"username": "john@example.com",
"password": "secret"
}]
)
```
```bash cURL theme={null}
curl -X POST "https://api.anchorbrowser.io/v1/identities" \
-H "anchor-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"source": "https://linkedin.com",
"name": "John Doe",
"metadata": {
"department": "Engineering",
"role": "admin"
},
"credentials": [{
"type": "username_password",
"username": "john@example.com",
"password": "secret"
}]
}'
```
### Filtering Identities by Metadata
You can filter identities by metadata when listing them for an application:
```javascript node.js theme={null}
const identities = await anchorClient.applications.listIdentities(app.id, {
metadata: JSON.stringify({ department: 'Engineering' })
});
```
```python python theme={null}
identities = anchor_client.applications.list_identities(
application_id="4d0cfce6-cf86-43eb-9955-1332941dbbb6",
metadata=json.dumps({"department": "Engineering"})
)
```
```bash cURL theme={null}
curl -G "https://api.anchorbrowser.io/v1/applications/{app_id}/identities" \
--data-urlencode 'metadata={"department":"Engineering"}' \
-H "anchor-api-key: YOUR_API_KEY"
```
## Related
Let end users connect accounts from your product
Alternative approach using browser profiles
Applications and Identities endpoints