> For the complete documentation index, see [llms.txt](https://help.blotato.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.blotato.com/start-with-an-ai-agent/agents.md).

# Start here

Route a customer task or support question by surface, client, platform, artifact, and stage.

This is the entry point for customer assistants executing tasks and support agents answering questions. Website instructions describe actions an agent performs or explains to the customer. They are not interchangeable with REST or MCP calls.

## Route the request

Use context already supplied. Ask only for a missing fact which changes the next action.

1. Identify the requested outcome and whether you are executing it or explaining it.
2. Identify the surface: Blotato website, MCP, REST, n8n, or Make.
3. Identify the client, destination platform, and failing stage when relevant.
4. Identify the artifact: workflow JSON, local media file, hosted media URL, generated visual, draft, scheduled item, or published post.
5. Check which provider owns the endpoint, key, or node. A Blotato key does not authenticate another provider.

For support, follow [Diagnose and answer](/start-with-an-ai-agent/agent-triage.md). Answer the customer's question before offering a tutorial. Preserve completed checks and human corrections.

## Website Modes and Current Navigation

Developer Platform is the website navigation for monitoring API, MCP, and automation work. Content Studio is the navigation for the in-app AI Agent and visual creation. Switching modes does not change the subscription, accounts, or content, and does not connect an external assistant.

* In-app AI Agent: Content Studio > Post editor. New draft inside the workspace creates a blank draft, not AI output.
* Connect or reconnect social destinations: Accounts > Connect account, or the existing account's actions menu > Reconnect. Accounts is no longer a Settings tab.
* Publication results: Posts > Scheduled, Published, or Failed. Select the platform and time window before concluding a post is missing.
* Recurring slots: Calendar > Weekly schedule, within the Weekly posting schedule section.
* Performance rankings: Analytics > Top posts, not the old Published > Top Performing tab.
* API activity and errors: Developer Platform > Logs.
* Instagram and Facebook comments and messages: Inbox > Posts or Chats.
* Inspiration, Prompts, and Viral AI Coach: Tools.
* Credentials and client setup: Settings > API. Language and reset controls: Settings > Profile. Credits and subscription: Settings > Billing.

Use [Website modes and navigation](/navigation.md) for the screen-by-screen map and screenshots. The draft's Photo / Video control uploads existing media. Do not send users there to find the old image-generation menu.

## Connect and verify

1. Select the exact client in [MCP setup](/start-with-an-ai-agent/mcp/setup.md), or use the [REST quickstart](/start-with-an-ai-agent/start.md) for HTTP, n8n, or Make.
2. Confirm paid access. Generating an API key during a trial starts billing. Read [authentication](/api-and-mcp-concepts/authentication.md) before generating a key.
3. Call `blotato_list_accounts`, or `GET /v2/users/me/accounts`.
4. Select the requested social destination from the response. Connect missing accounts in [Accounts](https://my.blotato.com/accounts).

An empty list proves authentication, not publishing readiness. A saved connector, OAuth redirect, or “Connected” label alone does not prove a tool call works. ChatGPT.com and Codex have separate setup paths.

## Choose a task

| Outcome                                                 | Execute through API or MCP                                                | Explain website actions                                                            |
| ------------------------------------------------------- | ------------------------------------------------------------------------- | ---------------------------------------------------------------------------------- |
| Connect a social destination                            | [IDs and destinations](/api-and-mcp-concepts/accounts-and-identifiers.md) | [Social account settings](/settings/social-accounts.md)                            |
| Publish or cross-post                                   | [Publish and verify](/agent-workflows/agent-workflows/publish.md)         | [Create posts](/web-app-getting-started/posts.md)                                  |
| Schedule, reschedule, or cancel                         | [Manage schedules](/agent-workflows/agent-workflows/schedule.md)          | [Calendar](/web-app-getting-started/calendar.md)                                   |
| Upload or use existing media                            | [Upload and prepare media](/agent-workflows/agent-workflows/upload.md)    | [Attach media](/support/faqs.md#how-do-i-create-an-instagram-reel-or-tiktok-video) |
| Generate images, carousels, or video                    | [Generate and verify](/agent-workflows/agent-workflows/visuals.md)        | [Video templates](/web-app-features/videos.md)                                     |
| Extract or repurpose source content                     | [Extract a source](/agent-workflows/agent-workflows/sources.md)           | [Context](/web-app-getting-started/sources.md)                                     |
| Inspect performance                                     | [Analytics](/agent-workflows/agent-workflows/analytics.md)                | [Checkpoints and pending data](/web-app-features/analytics.md)                     |
| Read or reply to Instagram and Facebook comments        | [Comments](/agent-workflows/agent-workflows/comments.md)                  | [Inbox](/web-app-features/inbox.md)                                                |
| Read or reply to Instagram and Facebook direct messages | [Messages](/agent-workflows/agent-workflows/messages.md)                  | [Inbox](/web-app-features/inbox.md)                                                |
| Set up keyword-triggered replies                        | [DM automations](/agent-workflows/agent-workflows/automations.md)         | [DM automation guide](/support/dm-automations.md)                                  |
| Check balance or purchase credits                       | [Credits](/agent-workflows/agent-workflows/credits.md)                    | [Billing and credits](/settings/billing-and-credits.md)                            |
| Diagnose a failure                                      | [Errors and retries](/api-and-mcp-concepts/error-handling.md)             | [Support diagnosis](/start-with-an-ai-agent/agent-triage.md)                       |

## Instructions for the agent

1. Use only available tools and the selected operation's schema. [REST and MCP](/api-and-mcp-concepts/rest-vs-mcp.md) differ in fields and operation coverage.
2. Retrieve current IDs. After reconnection, re-list accounts. Never substitute a social-platform post ID, submission ID, or schedule ID for another ID type.
3. Check [platform requirements](/platform-publishing/platforms.md) and [media requirements](/rest-api-reference/publish-post/media.md) before a write.
4. Confirm the requested destination and authorization. Connection tests must not publish, send messages, purchase credits, or activate automations.
5. Record returned IDs. Use [operation-specific completion rules](/api-and-mcp-concepts/async-jobs-and-polling.md), including error fields inside successful HTTP or MCP responses.
6. Inspect existing work after an uncertain write. Do not blindly repeat a create request.
7. Report the verified status, scheduled time, or returned media/public URL. Explain what remains unresolved.

## Documentation for retrieval

Start with the task page, then retrieve only the contract or diagnosis needed.

| Resource                                                          | Use                                                                      |
| ----------------------------------------------------------------- | ------------------------------------------------------------------------ |
| [Search and read help](/start-with-an-ai-agent/docs-search.md)    | Optional public documentation MCP lookup, separate from account access   |
| [MCP tools](/start-with-an-ai-agent/mcp/tools.md)                 | Available tools and behavior                                             |
| [MCP input index](/start-with-an-ai-agent/mcp/inputs.md)          | Required and optional tool arguments                                     |
| [API reference for agents](/start-with-an-ai-agent/llm.md)        | Generated setup, contracts, examples, and completion rules in 1 document |
| [Documentation index](https://help.blotato.com/llms.txt)          | Find a page outside this task index                                      |
| [Full documentation text](https://help.blotato.com/llms-full.txt) | Whole-corpus retrieval when required                                     |
| [OpenAPI JSON](https://backend.blotato.com/openapi.json)          | Published REST schemas                                                   |

Reading documentation does not install a server, authorize an account, or execute a task.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://help.blotato.com/start-with-an-ai-agent/agents.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.