### Install Dot CLI on self-hosted instance

Source: https://docs.getdot.ai/developers/cli.md

Install the Dot CLI from a self-hosted Dot instance using curl.

```bash
curl https://your-dot-instance.com/install | sh
```

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

### Install Dot CLI on macOS/Linux (US and EU regions)

Source: https://docs.getdot.ai/developers/cli.md

Install the Dot CLI on macOS or Linux using curl. Use the URL matching your Dot region: app.getdot.ai for US, eu.getdot.ai for EU.

```bash
# US region
curl https://app.getdot.ai/install | sh

# EU region
curl https://eu.getdot.ai/install | sh
```

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

### POST /api/agentic request example

Source: https://docs.getdot.ai/developers/api/commonly-used-endpoints

Example HTTP request to ask Dot a question. Requires an API key in the X-API-KEY header and a JSON body with messages, chat_id, and optional skip_check and extra fields. The response is 'No content'.

```http
POST /api/agentic HTTP/1.1
X-API-KEY: YOUR_API_KEY
Content-Type: application/json
Accept: */*
Content-Length: 104

{
  "messages": [
    {
      "ANY_ADDITIONAL_PROPERTY": "anything"
    }
  ],
  "chat_id": "text",
  "skip_check": false,
  "extra": "text"
}
```

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

### GET /api/export_history request example

Source: https://docs.getdot.ai/developers/api/commonly-used-endpoints

Example HTTP request to the export_history endpoint. Requires an admin API token passed in the X-API-KEY header. The response is a 200 with no content in this example.

```HTTP
GET /api/export_history HTTP/1.1
X-API-KEY: YOUR_API_KEY
Accept: */*
```

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

### Direct HTTP/HTTPS connection example

Source: https://docs.getdot.ai/integrations/databases/clickhouse.md

Shows the connection details for the recommended direct HTTP/HTTPS method, including host, port, username, password, and database.

```text
Host: clickhouse.company.com
Port: 8443
Username: dot_user
Password: your_secure_password
Database: analytics_db
```

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

### Create a User and Assign Privileges

Source: https://docs.getdot.ai/integrations/databases/mysql-mariadb

Execute these SQL commands as a superuser to create a user 'dot_user' and a role 'dot_role', then assign the role to the user. Replace '<something secret>' with a strong password.

```sql
CREATE USER 'dot_user'@'%' IDENTIFIED BY '<something secret>';

CREATE ROLE 'dot_role';

GRANT 'dot_role' TO 'dot_user'@'%';
```

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

### Agent Operating Principles Example

Source: https://docs.getdot.ai/train-dot/model/notes

An example of agent operating principles, written as instructions with strong language like 'Always' and 'Never' to guide Dot's behavior in specific situations.

```text
Always clarify with the user if they want live or booked revenue when they ask about revenue.
Always by default filter out internal users (IS_INTERNAL = FALSE).
Always use the table fct_arr_consolidated when this table is enough to answer questions about ARR.
Never answer questions about a/b tests and rather refer them to https://experiments.internal.com
```

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

### curl command for GET /api/chat_history_export

Source: https://docs.getdot.ai/developers/api/commonly-used-endpoints.md

Example curl command to export chat history as a CSV file. It uses the X-API-KEY header with an admin API token and passes start_date, end_date, and tz parameters. The example uses the EU endpoint https://eu.getdot.ai and saves the output to history.csv.

```bash
# EU: https://eu.getdot.ai
curl -H "X-API-KEY: YOUR_API_TOKEN" -o history.csv \
  "https://app.getdot.ai/api/chat_history_export?start_date=2026-10-01&end_date=2026-10-31&tz=Europe/Berlin"
```

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

### Full example URL for embedding Dot with UI parameters

Source: https://docs.getdot.ai/developers/embed.md

A complete URL example that combines multiple optional parameters to customize the embedded Dot UI, such as hiding side navigation, help, share button, explanation, and setting a primary color and chat placeholder text. Use this as a template for constructing your own embed URLs.

```url
https://eu.getdot.ai?hideSideNavigation=true&hideHelp=true&hideShareButton=true&hideExplanation=true&minimizeProgess=True&primaryColor=%23c700c7&chatPlaceHolderText=Ask%20AI
```

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

### GET /api/get_users - Get All Users

Source: https://docs.getdot.ai/developers/api/commonly-used-endpoints

Retrieves all users. Requires X-API-KEY header. Returns 200 with no content.

```HTTP
GET /api/get_users HTTP/1.1
X-API-KEY: YOUR_API_KEY
Accept: */*
```

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

### Log in to Dot CLI

Source: https://docs.getdot.ai/developers/cli.md

Log in to Dot CLI. Opens a browser to sign in and saves the token locally. Use --device to print a sign-in link, then run dot login --poll <code> with the printed code. Use --browser to force the browser flow.

```bash
dot login
```

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

### Preview and run an evaluation

Source: https://docs.getdot.ai/developers/api/evaluations-in-ci.md

Runs a dry-run preview that starts no agent and spends no evaluation credits, then runs the evaluation and saves JSON and JUnit reports. The run grades the file's questions without changing the saved definition.

```bash
dot eval run suite.yaml --target production --dry-run
dot eval run suite.yaml --target production \
  --output artifacts/results.json --junit artifacts/results.xml
```

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

### GET /api/c2/{chat_id}

Source: https://docs.getdot.ai/developers/api/commonly-used-endpoints

Reads one whole conversation by chat ID. Use include_all=true to get the full conversation.

```APIDOC
## GET /api/c2/{chat_id}

### Description
Reads one whole conversation by chat ID.

### Method
GET

### Endpoint
/api/c2/{chat_id}

### Parameters
#### Path Parameters
- **chat_id** (string) - Required - The ID of the chat to retrieve.

#### Query Parameters
- **include_all** (boolean) - Optional - Set to `true` to include the full conversation.

### Response
#### Success Response (200)
- **conversation** - The full conversation data.
```

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

### dot env --help

Source: https://docs.getdot.ai/developers/cli.md

Displays help for additional environment verbs: rename, delete, unset, conflicts, target, and sync-target.

```bash
dot env --help
```

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

### Organization Notes Example

Source: https://docs.getdot.ai/train-dot/model/notes

A short example of an organization note providing background knowledge about a company. The source notes that this is much shorter than typical background notes.

```text
TinyTrucks is a children's toy company. 
We manufature educational and entertaining toy trucks.
We sell directly them online to parents in Europe and the US.
```

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

### Metric Glossary Example

Source: https://docs.getdot.ai/train-dot/model/notes.md

Example of a metric glossary definition that Dot will use when answering queries. It includes a header instruction and SQL queries for ARR and NRR metrics.

```markdown
This is our metric glossary. 
Always use the definitions specified here to write queries

## ARR: Annual Recurring Revenue
```sql
SELECT sum(contract_value)
FROM opportunities_table
WHERE status = 'closed'
```

## NRR: Net Revenue Retention
...
```

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

### Typical agent recipe: fix documentation for a dbt model

Source: https://docs.getdot.ai/train-dot/environments.md

A workflow for fixing documentation in a dbt model: create and use an environment, set a target schema, build the model, sync the target, verify with `dot ask`, then merge and delete the environment. Assumes a connection ID and a dbt project.

```bash
dot env create "fix-customer-tier" && dot env use fix-customer-tier
dot env target set fix-customer-tier <connection-id> --schema dbt_dev_alice
dbt build --select dim_customers     # build the model into the dev schema
dot env sync-target fix-customer-tier
dot ask "Which columns does dim_customers have?"   # verify Dot sees the dev version
dot env merge fix-customer-tier --confirm --delete-after
```

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

### Python: Main function to create Notion page

Source: https://docs.getdot.ai/train-dot/custom-skills

Main function that creates a Notion page with properties, checks required environment variables, and prints success or error messages. Requires NOTION_DATABASE_ID and NOTION_API_TOKEN. Exits with error if missing.

```python
def main(name, description, status_name, priority, due_date):
    """Main function for testing."""
    setup_logging()
    
    # Check for required environment variables
    db_id = os.getenv("NOTION_DATABASE_ID")
    if not db_id:
        print("❌ NOTION_DATABASE_ID not set")
        print("Set it with: export NOTION_DATABASE_ID='your_database_id'")
        sys.exit(1)
    
    token = os.getenv("NOTION_API_TOKEN")
    if not token:
        print("❌ NOTION_API_TOKEN not set")
        print("Set it with: export NOTION_API_TOKEN='your_token'")
        sys.exit(1)
    
    try:
        # Create page with only Name property (since that's what your database has)
        notion = NotionWrapper()
        properties = {
            "Name": title_prop(f"{name}"),
            "Description": rich_text(description),
            "Status": status(status_name),
            "Due Date": date(due_date),
            "Priority": select(priority)
        }
        page_id = notion.create_page(db_id, properties)
        
        # Generate URLs
        page_url = get_page_url(page_id)
        database_url = get_database_url(db_id)
        
        print("\n🎉 Success! Created Notion page:")
        print(f"📝 Page ID: {page_id}")
        print(f"🔗 Page URL: {page_url}")
        print(f"📊 Database URL: {database_url}")
        print("\nYou can click these URLs to view in Notion!")
        
        return {
            "page_id": page_id,
            "page_url": page_url,
            "database_id": db_id,
            "database_url": database_url
        }
        
    except NotionWrapperError as e:
        print(f"❌ Error: {e}")
        sys.exit(1)
```