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

# MCP tools

> Browse the generated KonbiniAPI MCP tool reference. One typed tool exists for each supported Instagram, TikTok, X, and Reddit endpoint.

KonbiniAPI's MCP server exposes one typed tool per supported endpoint. This
reference stays aligned with the current MCP tool catalog. Agents can call
the matching tool directly from their MCP client.

## Shared optional inputs

All endpoint tools expose endpoint parameters as top-level typed inputs. They
also share the same projection controls so you can keep payloads small,
save tokens, and keep results deterministic.

| Parameter           | Type       | Description                                                                                                          |
| ------------------- | ---------- | -------------------------------------------------------------------------------------------------------------------- |
| `projection_preset` | `string`   | Optional. Defaults to `minimal`. Allowed values: `full`, `minimal`, `identity`, `engagement`, `content`.             |
| `data_fields`       | `string[]` | Optional. Adds top-level keys from `data` on top of the selected `projection_preset`.                                |
| `item_fields`       | `string[]` | Optional. Only on collection tools. Adds keys from `data.orderedItems[]` on top of the selected `projection_preset`. |

## Projection options

Use projections to keep tool results focused on the data your agent actually
needs. Smaller payloads mean fewer tokens consumed per call.

| Preset       | Best for                      | What it keeps                                                                                                                                                           |
| ------------ | ----------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `full`       | Full payload access           | Returns the full API response without projection.                                                                                                                       |
| `minimal`    | Default lookups               | Keeps a small set of commonly useful identity fields. On collection tools, it also keeps paging fields such as `cursor`, `nextCursor`, `itemCount`, and `orderedItems`. |
| `identity`   | Profiles and account metadata | Focuses on identity and profile fields such as names, bios, verification state, and audience counts.                                                                    |
| `engagement` | Metrics and comparisons       | Focuses on counts such as views, likes, comments, shares, reposts, saves, and follower totals.                                                                          |
| `content`    | Media and captions            | Focuses on content fields such as text, media attachments, tags, duration, authorship, and language.                                                                    |

### Adding fields on top

* Use `data_fields` to add top-level keys from `data` on top of the selected preset.
* Use `item_fields` on collection tools to add keys from each item in `data.orderedItems[]`.
* If you know exactly what you need, start with `minimal` and add only the extra fields required for the task.

## Response shape

Successful tool calls keep the API envelope under `data` and add MCP credit
metadata when available:

```json theme={null}
{
  "data": {},
  "metadata": {
    "creditsUsed": 1,
    "creditsRemaining": 4999
  }
}
```

## Browse the catalog

Use the sidebar navigation to browse all available tools by platform. Each tool
has its own page with parameters, example input, and available projected fields.