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

# MCP argument examples

> Valid argument shapes for discovery, reporting, campaign validation, memory, and launch polling.

These examples show argument objects for named MCP tools. Send them through your authenticated client. Replace example UUIDs with IDs from discovery reads and generate a fresh request UUID for a new operation.

The read examples do not change advertising objects. `memory_remember` stores a fact; use it only with consent. The campaign example sets `dry_run: true`.

## List accessible accounts

**Tool:** `ad_accounts_list`

Read account IDs, currencies, and time zones before choosing a target.

```json theme={null}
{
  "limit": 25
}
```

## Read daily campaign metrics

**Tool:** `insights_query`

Replace the account UUID and dates. Call `insights_capabilities` first. Check `complete` and `next_cursor` in the result.

```json theme={null}
{
  "ad_account_ids": [
    "00000000-0000-4000-8000-000000000001"
  ],
  "date_from": "2026-10-01",
  "date_to": "2026-10-07",
  "metrics": [
    "spend",
    "impressions",
    "clicks"
  ],
  "entity_level": "campaign",
  "report_spec": "base_daily",
  "grain": "daily",
  "limit": 25
}
```

## Rank purchase ROAS

**Tool:** `insights_rank`

Choose the purchase event and attribution window that match your analysis. The result distinguishes numeric extremes and includes data-through information.

```json theme={null}
{
  "ad_account_ids": [
    "00000000-0000-4000-8000-000000000001"
  ],
  "date_from": "2026-10-01",
  "date_to": "2026-10-07",
  "entity_level": "campaign",
  "rank_by": "roas",
  "action_type": "omni_purchase",
  "attribution_window": "7d_click",
  "direction": "both",
  "rank_scope": "global",
  "min_spend": 100,
  "limit": 5
}
```

## Validate a paused campaign

**Tool:** `campaigns_create`

This example validates the inputs without creating a campaign. Check `actions_capabilities` first. Review the result before a live request; use a new `request_id` when changing `dry_run` to `false`.

```json theme={null}
{
  "payload": {
    "request_id": "00000000-0000-4000-8000-000000000002",
    "ad_account_id": "00000000-0000-4000-8000-000000000001",
    "name": "October traffic test",
    "objective": "OUTCOME_TRAFFIC",
    "special_ad_categories": [],
    "status": "PAUSED",
    "dry_run": true
  }
}
```

## Save a fact after explicit consent

**Tool:** `memory_remember`

Use this only after the user asks you to save the fact or approves that suggestion. Replace the brand UUID and generate a request UUID. Inspect the returned persistence state.

```json theme={null}
{
  "payload": {
    "request_id": "00000000-0000-4000-8000-000000000002",
    "fact": "Use British English for this brand's campaign copy.",
    "scope": {
      "kind": "brand",
      "target_id": "00000000-0000-4000-8000-000000000003"
    },
    "consent": "explicit_user_request"
  }
}
```

## Find failed tool activity

**Tool:** `activity_list`

Filter by the exact tool name and failed status. Follow `next_cursor` with the same filters to read the complete set.

```json theme={null}
{
  "payload": {
    "tool_name": "campaigns_create",
    "status": "failed",
    "limit": 25,
    "cursor": 0
  }
}
```

## Discover saved creatives

**Tool:** `launcher_creatives_list`

Inspect the returned version and approval information before selecting content.

```json theme={null}
{
  "limit": 25
}
```

## Poll an accepted launch

**Tool:** `launcher_launches_get`

Use the account and launch IDs from the submission. Use `launcher_run_operations` to check the outcome of each operation.

```json theme={null}
{
  "ad_account_id": "00000000-0000-4000-8000-000000000001",
  "launch_id": "00000000-0000-4000-8000-000000000004"
}
```


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.