# merge_chat_changes

Source: https://modulify.ai/docs/api/chat/merge-chat-changes

Tries again to merge the unmerged changes of a chat into the site.

- Title: Merge the unmerged changes of a chat
- Scope: `chat:write`
- Access: Makes changes
- Endpoint: `POST /v1/merge_chat_changes`

Each chat works in its own copy of the site. When the end of a turn cannot merge its changes automatically, for example because another chat changed the same lines, they are kept as unmerged changes, which `list_chats` shows as `pending`. This tries the merge again. See [Changes not merged](https://modulify.ai/docs/build/multiple-chats#changes-not-merged).

The response carries the outcome in `integration`, where `state` reads:

| `state` | What it means |
| --- | --- |
| `landed` | The changes are on the site now |
| `no-changes` | Nothing was left to merge |
| `conflicted`, `pending` or `rejected` | They still could not be merged and stay unmerged. Send a message to that chat asking the agent to resolve the conflicts |

`integration` also lists the files that conflict in `conflicts`, each with its `path` and `kind`.

It is refused with a `409` while a generation runs in that chat or its changes are already being merged or discarded (`Wait for this chat to finish first!`), while the site finishes a restore or move or its preview moves to a new [environment](https://modulify.ai/docs/editor/environment) or a bigger machine (`Changes cannot be saved to this site while it finishes a restore or move. Try again in a moment!`) and when the chat has no unmerged changes (`This chat has no unmerged changes!`). It is also refused while the site is locked, like every tool that changes a locked site, and while its GitHub repository blocks editing, with the same message `send_message` gives.

Two refusals are safe to try again in a moment:

| Status | When | The message |
| --- | --- | --- |
| `409` | Another chat's changes are being merged into the site right now | `Another chat is merging changes right now. Try again in a moment!` |
| `503` | The site or its repository could not be reached to finish the merge | `The site could not be reached to finish this. Try again in a moment!` |

## Request

Call it with a `POST` to `https://api.modulify.ai/v1/merge_chat_changes`, sending the inputs below as a JSON object. The token needs the `chat:write` scope.

It makes changes, so send an [Idempotency-Key](https://modulify.ai/docs/api/idempotency) header whenever you might retry it. A retry with the same key gets the first answer back instead of running again.

```bash
curl -X POST https://api.modulify.ai/v1/merge_chat_changes \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"projectId":"PROJECT_ID"}'
```

Over MCP, the same method is the [merge_chat_changes tool](https://modulify.ai/docs/mcp/chat/merge-chat-changes).

## Inputs

| Input | Type | Required | Description |
| --- | --- | --- | --- |
| `projectId` | string | Yes | The site id, from `list_sites` or `get_site`. |
| `chatId` | string | No | The chat to merge: `main` for the chat every site comes with, or the `chatKey` of another chat from `list_chats`. Defaults to `main`. |

## Response

Every call answers with the [JSON envelope](https://modulify.ai/docs/api/requests-and-responses#the-response) of `success`, `message`, `data`, `code` and `version`. `data` holds the result described above, and on a method that returns a total, `count` carries it. The [response headers](https://modulify.ai/docs/api/requests-and-responses#headers-on-every-method-call) carry the call's `X-Request-Id` and what is left of your per-minute budget in `X-RateLimit-Limit`, `X-RateLimit-Remaining` and `X-RateLimit-Reset`. [Errors](https://modulify.ai/docs/api/errors) explains every status code a call can answer with.