- Brand
- Flowera AI
- Category
- AI
- Primary Subcategory
- AI Chat & Messaging Agents
Integration details
Description
Flowera builds agentic AI automations for WhatsApp and Instagram: AI agents that answer your customers' messages, qualify leads, book appointments and hand over to your team when needed. Connect your Flowera account and design, change and test those agents (AgentFlows) just by describing what you want. You can: - Build WhatsApp and Instagram agents from a plain-language description, or change existing ones: add, connect and configure steps, choose the AI model, set up webhooks. - List, open and explain your flows step by step, validate them before going live, compare versions and restore an older one. - See your WhatsApp and Instagram automations (scheduled and batch campaigns) and how they are doing. - Work with the rest of Flowera too: web chat agents, captured leads and stored files. You stay in control. On connecting you pick one workspace and approve each permission separately. By default the app can read and build flows and see automations and files; running a flow (which can message real customers), reading customer conversations, seeing connected accounts and sending or changing automations are separate permissions that are requested only when needed. Deleting a flow is not possible from ChatGPT; it stays in the Flowera app. Requires a Flowera account (portal.flowera.ai).
- Integration type
- Plugin
- Verification status
- Not applicable
- Platform
- ChatGPT
- Primary Subcategory
- AI Chat & Messaging Agents
- Secondary Subcategories
- None listed
- Brand
- Flowera AI
- Access
- Account required
- First tracked
- 2026-11-21
- Tool count
- 163
- Market
- US
The Primary Subcategory used for this profile’s headline score.
Other Subcategories where the Integration is listed.
Not discoverable in ChatGPT yet
ChatGPT now suggests Plugins on its own when they match a user's request, but not every Plugin is being surfaced yet.We haven’t seen Flowera suggested yet. Sign up and we’ll track it, and your Discoverability Score will appear here once ChatGPT starts surfacing it.
What discovery looks like

Get alerts for Flowera
Get updates when Flowera’s Discoverability Score or category rank changes.
Competing in ChatGPT AI Chat & Messaging Agents
View Category163 tools agents can invoke
Put a saved automation live: from now on it fires on its own schedule with nobody watching. Refuses, with a list of blockers, if it is not ready — a start time in the past, a recipient file that no longer exists, a CSV with no column mapped, a lead filter matching nobody. Also the way to restart a paused automation that you have just edited. The call must restate what the automation runs, as read from the automation: its flow id, its execution mode and its recipient source; a call whose values differ is refused. Note: from now on it runs by itself and can message your customers. Starts or sends only a flow whose every step was made with this app's step tools; it refuses any other flow.
activate_automation
Adds an Agent step to one of your saved flows. An AI step that reasons over the conversation and answers. The step has no tool steps of its own; its built-in model tools, such as web search, can be set. Keep showOutputInChat off for an AI step: it only shows the output in the message view in the app and never sends anything. The AI model is chosen in a separate step. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_agent_node
Adds a Condition Agent step to one of your saved flows. Utilize an agent to split flows based on dynamic conditions. Each scenario is one output branch, in order, and there is no automatic else branch: add a last scenario such as "other", or messages that match none are dropped. Its input must be {{$question}}, the default. The AI model is chosen in a separate step. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_condition_agent_node
Adds a Condition step to one of your saved flows. Split flows based on If Else conditions. Each condition row is one output branch, in order, and a last Else branch takes everything else. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_condition_node
Adds a Custom Function step to one of your saved flows. Execute custom function. It runs your JavaScript; its return value is the step output. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_custom_function_node
Adds a Date and Time tool step to one of your saved flows. It is a Tool step bound to CurrentDateTime: Get todays day, date and time. It returns the current date and time. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_current_date_time_tool_node
Adds a Direct Reply step to one of your saved flows. Reply in the chat that ran the flow (test chat, web chat, API). It answers in the Flowera chat only: a WhatsApp, Instagram or Facebook customer never sees it. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_direct_reply_node
Adds an Execute Flow step to one of your saved flows. Execute another flow. It runs another flow of your workspace with the given input and returns its answer. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_execute_flow_node
Adds a Facebook message step to one of your saved flows. Reply on the account the flow is triggered on: a Direct Reply step never reaches the customer. Keep showOutputInChat on: it only shows the send in the message view. It sends a text message to the Page customer who wrote. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_facebook_send_message_step
Adds a Filter step to one of your saved flows. Filter data with condition groups and AND/OR logic. It has two outputs: True when the condition groups match, False otherwise. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_filter_node
Adds a Gmail email step to one of your saved flows. It works with a Google account connected in the Flowera app (credentialId); it can be set later. It sends an email from the connected Gmail account. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_gmail_send_email_step
Adds a Google Calendar availability check step to one of your saved flows. It works with a Google account connected in the Flowera app (credentialId); it can be set later. It tells whether calendars are free or busy in a time window. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_google_calendar_check_availability_step
Adds a Google Calendar event step to one of your saved flows. It works with a Google account connected in the Flowera app (credentialId); it can be set later. It creates an event in a calendar; attendees given are invited by Google. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_google_calendar_create_event_step
Adds a Google Search tool step to one of your saved flows. It is a Tool step bound to Google Custom Search: Wrapper around Google Custom Search API - a real-time API to access Google search results. It needs a Google Custom Search credential (credentialId). Returns the new step's id; with connectFrom it is also connected after an existing step.
add_google_custom_search_tool_node
Adds a Google Sheets read step to one of your saved flows. It works with a Google account connected in the Flowera app (credentialId); it can be set later. It reads the values of a range of a sheet. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_google_sheets_read_rows_step
Adds a Google Sheets row append step to one of your saved flows. It works with a Google account connected in the Flowera app (credentialId); it can be set later. It adds rows after the last filled row of a sheet. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_google_sheets_append_rows_step
Adds a Google Sheets update step to one of your saved flows. It works with a Google account connected in the Flowera app (credentialId); it can be set later. It writes values into a range of a sheet, replacing what is there. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_google_sheets_update_rows_step
Adds an HTTP step to one of your saved flows. Send a HTTP request. It calls an external URL. Give body as text for a raw or JSON body, or formBody rows for a form. A key or token belongs in a credential, never in a header value. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_http_node
Adds a Human Input step to one of your saved flows. Request human input, approval or rejection during execution. It pauses the flow until a person proceeds or rejects: two outputs, Proceed and Reject. The AI model is chosen in a separate step. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_human_input_node
Adds an Instagram comment reply step to one of your saved flows. Reply on the account the flow is triggered on: a Direct Reply step never reaches the customer. Keep showOutputInChat on: it only shows the send in the message view. A comment and a direct message are different events: branch on the trigger event type, and use the post caption for a comment. The trigger name must contain "instagram" (or the word "ig", as in "ig-dm"). It answers in public, under the customer's comment. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_instagram_reply_to_comment_step
Adds an Instagram image step to one of your saved flows. Reply on the account the flow is triggered on: a Direct Reply step never reaches the customer. Keep showOutputInChat on: it only shows the send in the message view. A comment and a direct message are different events: branch on the trigger event type, and use the post caption for a comment. The trigger name must contain "instagram" (or the word "ig", as in "ig-dm"). It sends an image, by its public link, in a direct message. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_instagram_send_image_step
Adds an Instagram message step to one of your saved flows. Reply on the account the flow is triggered on: a Direct Reply step never reaches the customer. Keep showOutputInChat on: it only shows the send in the message view. A comment and a direct message are different events: branch on the trigger event type, and use the post caption for a comment. The trigger name must contain "instagram" (or the word "ig", as in "ig-dm"). It sends a direct message to the customer who wrote or commented. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_instagram_send_message_step
Adds an Instagram private reply step to one of your saved flows. Reply on the account the flow is triggered on: a Direct Reply step never reaches the customer. Keep showOutputInChat on: it only shows the send in the message view. A comment and a direct message are different events: branch on the trigger event type, and use the post caption for a comment. The trigger name must contain "instagram" (or the word "ig", as in "ig-dm"). It answers a comment with a direct message to its author, once per comment. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_instagram_private_reply_step
Adds an Iteration step to one of your saved flows. Execute the nodes within the iteration block through N iterations. It runs the steps inside it once per item of its input array; add those steps with parentNode set to its id. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_iteration_node
Adds a JSON Path tool step to one of your saved flows. It is a Tool step bound to JSON Path Extractor: Extract values from JSON using path expressions. It reads one value out of JSON by its path. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_json_path_extractor_tool_node
Adds an LLM step to one of your saved flows. Large language models to analyze user-provided inputs and generate responses. Keep showOutputInChat off for an AI step: it only shows the output in the message view in the app and never sends anything. The AI model is chosen in a separate step. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_llm_node
Adds a Loop step to one of your saved flows. Loop back to a previous node. It goes back to an earlier step, at most maxLoopCount times. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_loop_node
Adds a PDF tool step to one of your saved flows. It is a Tool step bound to PDF Generator: Render a PDF from an HTML template (or a registered templateId) + data, via the Flowera PDF platform. It makes a PDF from a saved template or from HTML. It needs a Flowera API credential (credentialId). Returns the new step's id; with connectFrom it is also connected after an existing step.
add_pdf_generator_tool_node
Adds a Retriever step to one of your saved flows. Retrieve information from vector database. It searches document stores and returns what it finds. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_retriever_node
Adds a Wait step to one of your saved flows. Pause execution for a specified duration. It pauses the flow for waitDuration, "HH:MM:SS", before the next step. Returns the new step's id; with connectFrom it is also connected after an existing step.
add_wait_node
Adds a WhatsApp buttons message step to one of your saved flows. Reply on the account the flow is triggered on: a Direct Reply step never reaches the customer. Keep showOutputInChat on: it only shows the send in the message view. The recipient is the sender of the trigger message; senderId may be a WhatsApp user id, senderPhone is the phone. The trigger name must contain "whatsapp". It sends a message with up to three reply buttons. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_whatsapp_send_buttons_step
Adds a WhatsApp form message step to one of your saved flows. Reply on the account the flow is triggered on: a Direct Reply step never reaches the customer. Keep showOutputInChat on: it only shows the send in the message view. The recipient is the sender of the trigger message; senderId may be a WhatsApp user id, senderPhone is the phone. The trigger name must contain "whatsapp". It sends a published WhatsApp form (Flow) with its call-to-action button. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_whatsapp_send_form_step
Adds a WhatsApp list message step to one of your saved flows. Reply on the account the flow is triggered on: a Direct Reply step never reaches the customer. Keep showOutputInChat on: it only shows the send in the message view. The recipient is the sender of the trigger message; senderId may be a WhatsApp user id, senderPhone is the phone. The trigger name must contain "whatsapp". It sends a message with a list of choices to pick from. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_whatsapp_send_list_step
Adds a WhatsApp media message step to one of your saved flows. Reply on the account the flow is triggered on: a Direct Reply step never reaches the customer. Keep showOutputInChat on: it only shows the send in the message view. The recipient is the sender of the trigger message; senderId may be a WhatsApp user id, senderPhone is the phone. The trigger name must contain "whatsapp". It sends an image, video, audio file or document, with an optional caption. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_whatsapp_send_media_step
Adds a WhatsApp template message step to one of your saved flows. In a campaign the recipient is the phone the list or the lead filter put in the flow state, e.g. {{ $flow.state.phone }}; in a reply it is senderPhone. Keep showOutputInChat on: it only shows the send in the message view. It sends an approved template message: the campaign send, and the only kind delivered more than 24 hours after the customer last wrote. The template and its values are set once this step is saved, as the result says. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_whatsapp_send_template_step
Adds a WhatsApp text message step to one of your saved flows. Reply on the account the flow is triggered on: a Direct Reply step never reaches the customer. Keep showOutputInChat on: it only shows the send in the message view. The recipient is the sender of the trigger message; senderId may be a WhatsApp user id, senderPhone is the phone. The trigger name must contain "whatsapp". It sends a text message to the customer who wrote. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_whatsapp_send_text_step
Adds a Zoho CRM note step to one of your saved flows. It works on the Zoho CRM connection given; the connection can be left empty and set later. Records are looked up and written with these steps, one per action, never by an Agent, and only once the details are complete or changed. It adds a note to a record. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_zoho_crm_add_note_step
Adds a Zoho CRM record creation step to one of your saved flows. It works on the Zoho CRM connection given; the connection can be left empty and set later. Records are looked up and written with these steps, one per action, never by an Agent, and only once the details are complete or changed. It creates a record, once the details are complete and a search found none. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_zoho_crm_create_record_step
Adds a Zoho CRM record lookup step to one of your saved flows. It works on the Zoho CRM connection given; the connection can be left empty and set later. Records are looked up and written with these steps, one per action, never by an Agent, and only once the details are complete or changed. It reads one record by its id. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_zoho_crm_get_record_step
Adds a Zoho CRM record update step to one of your saved flows. It works on the Zoho CRM connection given; the connection can be left empty and set later. Records are looked up and written with these steps, one per action, never by an Agent, and only once the details are complete or changed. It changes a record found by a search, keeping its id; it never overwrites a value the customer did not give. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_zoho_crm_update_record_step
Adds a Zoho CRM search by email step to one of your saved flows. It works on the Zoho CRM connection given; the connection can be left empty and set later. Records are looked up and written with these steps, one per action, never by an Agent, and only once the details are complete or changed. It looks a record up by email address. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_zoho_crm_search_by_email_step
Adds a Zoho CRM search by phone step to one of your saved flows. It works on the Zoho CRM connection given; the connection can be left empty and set later. Records are looked up and written with these steps, one per action, never by an Agent, and only once the details are complete or changed. It looks a record up by phone number, the first step before an update or a create. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_zoho_crm_search_by_phone_step
Adds a Zoho CRM search step to one of your saved flows. It works on the Zoho CRM connection given; the connection can be left empty and set later. Records are looked up and written with these steps, one per action, never by an Agent, and only once the details are complete or changed. It finds records whose fields match the criteria given. For a value the customer gave, such as a phone number or an email address, use the phone or email search step instead. Returns the new step's id and the names its arguments take; with connectFrom it is also connected after an existing step. It saves a step; nothing is sent or written now.
add_zoho_crm_search_step
End an automation permanently. It stops firing, every recipient still waiting is cancelled, and anything mid-send is aborted. This cannot be undone — the automation stays visible for its history but can never run again. Use pause_automation if you might want it back.
cancel_automation
Connect a WhatsApp / Instagram account to a flow in one call: its channel trigger AND every step that sends on that channel get it (one on another account is kept: keptSteps), saved as one version, validated, so no send step is missed and left unable to answer. `account`: an id from list_whatsapp_accounts / list_instagram_accounts. `argsToRead`: send steps whose arguments are still empty (their step tool lists them). `notes` names other flows already answering that account. From now on this flow answers every incoming message on that account.
connect_channel_account
Connect two nodes in an AgentFlow (add an edge from source to target). For multi-output nodes (Condition/HumanInput) pass `sourceOutput` (0-based) and optionally an `edgeLabel`. The correct source handle is read from the source node's real output anchors. A step added without a `position` moves beside the source on its first connection, so a Condition's branches stack in one column.
connect_nodes
Start connecting a WhatsApp Business number to this workspace. Returns a one-time link (valid 15 min) to a Flowera page where the user confirms the workspace, presses Connect and completes Meta's popup (Facebook login, consent, number verification) — the only steps Meta requires a person for. Flowera does the rest server-side. Show `url` to the user as a clickable link, then poll get_whatsapp_connection_status. Calling this again replaces your previous unused link.
connect_whatsapp
Creates a new flow in your workspace and returns its id. A new flow starts with only its Start step; the step tools add the rest, one step at a time.
create_agentflow
Create an automation — a saved schedule that runs an AgentFlow later, once or repeatedly. It is saved PAUSED and does not run until activate_automation, which also checks it is ready. Two shapes: executionMode "single" runs the flow once per fire; "fan-out" runs it once per recipient from a CSV file (upload_csv first) or a live lead filter (preview_lead_query first). The response includes `blockers` — anything that would stop activation — so you can fix it before arming.
create_automation
Create a WhatsApp message template and submit it to Meta review. Two steps: WITHOUT `confirm: true` it only checks it against the editor's and Meta's rules (errors with field paths), shows a text `preview` and the exact Meta `components`, and sends NOTHING — show that to the user. Then call again with `confirm: true`. Give `template` (the editor shape), or `preset` (order_confirmation, shipping_update, appointment_reminder, otp_code, promotion, product_carousel, blank) with `template` overrides (name and language at least), or raw Meta `components` + name/category/language for features beyond the editor. Media headers need upload_whatsapp_template_media first. Result is usually PENDING: poll get_whatsapp_template (review ≤24 h). Meta may re-categorize (e.g. UTILITY → MARKETING). Max 100 creations per account per hour.
create_whatsapp_template
Flowera ChatGPT Plugin FAQ
How the directory, categories and Discoverability Score work.
Read the methodologyHow do I improve Flowera's ChatGPT Plugin discoverability?
The levers are the listing surface agents actually read: names, descriptions, keywords, tool metadata, and registry health. Which lever matters depends on where discovery breaks, which is what continuous measurement shows.
What are Flowera alternatives on ChatGPT?
As of 2026-11-21, Flowera competes with 3Chat Customer Growth, Actually Asked, Aide, Ailways, aiwa, Algomo, Alhena AI, AssistantLabs Business Agent and 68 more in ChatGPT AI Chat & Messaging Agents, ranked by public Discoverability Score.
Where is this profile measured?
This profile uses the geography attached to the latest public registry snapshot: US. Locale tags are intentionally omitted.