Backlog

Next up

In Progress

Done

Main Roadmap

Lofty (formerly Chime) is a popular CRM for real estate.

In Review

Lofty CRM Integration

Robert Clements

Voice Input / Output

@seanp_ai

Agents: Tool use - Browser

I hope that the credits from Agenticflow can also be used for PixelML's features, even if it is converted at a fixed rate. This way, the full potential of this platform can be maximized.

Agenticflow's Credits can be used with Pixel ML's models.

DingChing Hong

Enable direct integration between AgenticFlow and Swipe One CRM to centralize customer data management and streamline workflow automation. This feature will allow users to automatically sync contacts, leads, and customer interactions from Swipe One CRM into AgenticFlow, providing a unified view of sales and customer engagement activities. Swipeone CRM API : <a target="_blank" rel="noopener noreferrer nofollow" class="link" href="https://docs.swipeone.com/en/collections/11679463-api">https://docs.swipeone.com/en/collections/11679463-api</a> By bridging the two platforms, users can create, update, and track CRM records, automate follow-ups, and trigger workflows based on CRM events—all within AgenticFlow. This integration aims to boost productivity, improve sales pipeline visibility, and enhance cross-team collaboration for organizations relying on both platforms.

Swipe One CRM Integration

vinaykumar

Voice Prompting (for lazy typing)

I have an agent working embed on my website, could we have it ?

Publish Agent to Facebook Fanpage

Bros Titan

I had a call with a T4 AS customer and realize there are use case that you need to get data behind the paywall. So current way of agenticflow is not design for this . In order to do this kind of automation, we need Agent to be able to use browser on it own. Would you be interested in this? Can you share the use case that you will do with it if we brought it to you? Few option available:<ul><li>BrowserBase</li><li>BrowserUse</li></ul>

AI Agent in AgenticFlow to use Browser

Automatically convert an agent chat session into a reusable, optimized workflow template.

Conversation‑to‑Workflow Conversion

Integrate AgenticFlow with Lancepilot WhatsApp CRM to enable unified customer communication and workflow automation. This feature allows users to send and receive WhatsApp messages, sync conversations, manage contacts, and trigger automated CRM workflows directly from within AgenticFlow. With this integration, users can centralize support and engagement, utilize rich messaging templates, and maintain real-time communication logs, streamlining both sales and customer service operations. Lancepilot API Documentation: <a target="_blank" rel="noopener noreferrer nofollow" class="link" href="https://documenter.getpostman.com/view/15730227/2sAYdimUKg">https://documenter.getpostman.com/view/15730227/2sAYdimUKg</a>

Lancepilot WhatsApp CRM Integration

Integrate with leading animation/video AI providers (Luma, Kling AI, etc.) for 3D and motion generation.

Planned

AI Animation & Video Animation

It would be great if we could add an MCP without connections.This would allow us to include:<ul><li>Code Nodes</li><li>Pieces in Embed Mode</li><li>Utility Nodes</li></ul>

Support MCP for nodes without connections

Hi, AgenticFlow is impressive, I would like to have more options to bring my OpenRouter key

OpenRouter BYOK support

Hi I want to enable my Agent to call Elevenlabs

ElevenLabs BYOK

We’d like to request a more detailed and interactive analytics dashboard within AgenticFlow, specifically tailored to track user engagement and workflow performance. Currently, metrics such as total users, total chats, and user activity return no data or are marked as unavailable.Proposed Enhancements:<ol><li>Real-Time Data Tracking – Show active users, agent usage frequency, and most engaged workflows over customizable timeframes (e.g., 7 days, 30 days).</li><li>Geo &amp; Channel Insights – Enable authorized location tracking to view regional rankings and distribution by publishing channels.</li><li>User Retention &amp; Depth Metrics – Visualize daily active users (DAU), user return rates, and average session lengths to help optimize engagement.</li><li>Agent Effectiveness Scores – Introduce a satisfaction or performance score based on task completion, user feedback, and repeat usage.</li><li>Exportable Reports – Allow users to download analytics reports (CSV, PDF) for external analysis and presentations.</li></ol>Why It Matters: This level of insight will empower AgenticFlow users to optimize workflows, measure ROI, and better understand which agents or automations drive the most value. <img data-width="100%" data-align="left" data-featurebase-content-key="67f8062dae8c6a55cdea186d/post/67fb2b1025d1191a187c206b/01962d20-2938-7fdb-a292-a24db6e58428/b64u-aW1hZ2UucG5n.png" src="https://67f8062dae8c6a55cdea186d.featurebase-attachments.com/c/post/67fb2b1025d1191a187c206b/01962d20-2938-7fdb-a292-a24db6e58428/b64u-aW1hZ2UucG5n.png?X-Amz-Expires=3600&amp;X-Amz-Date=20260220T230000Z&amp;X-Amz-Algorithm=AWS4-HMAC-SHA256&amp;X-Amz-Credential=DO801TYC4FCVNNEKURKM%2F20260220%2Ffra1%2Fs3%2Faws4_request&amp;X-Amz-SignedHeaders=host&amp;X-Amz-Signature=7080afa78b27b4f7799494421394c5e3f7722aef027d32e2f6273af7c8ea20fd" data-featurebase-content-filename="image.png">

Agent Analytics Dashboard

Add a native “approve/reject” step so agents can pause for manual validation before proceeding.

Built‑in Approval Workflow Node

Control flow allows us to create workflows that can make decisions, repeat tasks, and handle different scenarios intelligently.<ul><li>Sequential: For straightforward processes where steps always happen in the same order (this is the current workflow in AgenticFlow).</li><li>Conditional: When you need to make decisions or handle different cases differently</li><li>Loops: When you need to process multiple items or repeat an action multiple times</li><li>Error Handling: Always! Every workflow should consider what happens when things go wrong</li></ul>And even Advanced Control Flow Patterns like:<ul><li>Nested ConditionsConditions within conditions, like "If it's a weekday AND it's before noon, do X; otherwise do Y."</li><li>Loop ControlControlling loop execution with break (exit early) and continue (skip to next iteration).</li><li>Asynchronous FlowOperations that run independently and may complete in any order.</li></ul>

Control Flow: Sequential, Conditional, Loops

Alvin Alvelino

Direct integration to google sheet, google docs, etc..

Google drive direct integration

EUROCORSI

Overview: I would like to propose the addition of a folder feature within AgenticFlow to enhance the organization and management of workflows, agents, library items, and connections. As the number of workflows and agents grows, it becomes increasingly challenging to navigate and locate specific items efficiently. Implementing a folder structure would significantly improve user experience and productivity. Proposed Features:<ol><li>Folder Creation:<ul><li>Allow users to create custom folders to categorize workflows, agents, and library items based on their projects, themes, or any other criteria.</li></ul></li><li>Drag-and-Drop Functionality:<ul><li>Enable users to easily move workflows, agents, and library items into the appropriate folders using a simple drag-and-drop interface.</li></ul></li><li>Nested Folders:<ul><li>Support the creation of nested folders for more granular organization, allowing users to create subcategories within main folders.</li></ul></li><li>Connections Organization:<ul><li>Extend the folder feature to include connections, enabling users to group related connections together for better management.</li></ul></li><li>Search and Filter Options:<ul><li>Implement search and filter functionalities within folders to help users quickly locate specific workflows, agents, or library items.</li></ul></li></ol>Benefits:<ul><li>Improved organization and accessibility of workflows and agents.</li><li>Enhanced user experience, leading to increased productivity.</li><li>Simplified navigation for users managing multiple projects or complex workflows.</li></ul> Side note: this was written by the help of Pixi 😁

Folder Organization for Workflows, Agents, Library, and Connections

TLDR: Requesting an OpenAI-compatible endpoint for better integration with Straico models, which would improve agent connections, offer affordable access, and expand model options while reducing user costs.I would like to request an OpenAI compatible endpoint as an additional connection for building agents. This feature would be highly beneficial for several reasons, one of which being addressing the current issue of connecting Straico models to agents.By utilizing the OpenAI compatible endpoint API settings from Alter, which already has direct integration with Straico, this connection can be established. I frequently use this setup with various tools, allowing me access to both my Alter and Straico models.Alter offers a free tier or a very affordable BYOK yearly fee. I strongly recommend opting for this payment, as it provides a cost-effective entry point to Straico models through the OpenAI compatible endpoint.Implementing this feature would not only facilitate this specific connection but also expand the range of models available on AgenticFlow users in general. It would also help in reducing the additional costs that users might incur.

요약: 상담원 연결을 개선하고, 저렴한 액세스를 제공하며, 사용자 비용을 줄이면서 모델 옵션을 확장할 수 있는 Straico 모델과의 더 나은 통합을 위해 OpenAI 호환 엔드포인트를 요청합니다.상담원 구축을 위한 추가 연결로 OpenAI 호환 엔드포인트를 요청하고 싶습니다. 이 기능은 여러 가지 이유로 매우 유용할 것이며, 그 중 하나는 현재 Straico 모델을 상담원과 연결하는 문제를 해결하는 것입니다.이미 Straico와 직접 통합되어 있는 Alter의 OpenAI 호환 엔드포인트 API 설정을 활용하면 이 연결을 설정할 수 있습니다. 저는 이 설정을 다양한 도구와 함께 자주 사용하여 Alter와 Straico 모델 모두에 액세스할 수 있습니다.Alter는 무료 티어 또는 매우 저렴한 BYOK 연간 요금을 제공합니다. 이 결제 방식은 OpenAI 호환 엔드포인트를 통해 Straico 모델에 대한 비용 효율적인 진입점을 제공하므로 이 방식을 선택하는 것을 강력히 추천합니다.이 기능을 구현하면 이 특정 연결을 용이하게 할 뿐만 아니라 일반적으로 에이전틱플로우 사용자가 사용할 수 있는 모델의 범위를 확장할 수 있습니다. 또한 사용자에게 발생할 수 있는 추가 비용을 줄이는 데에도 도움이 될 것입니다.

OpenAI 호환 엔드포인트 연결(에이전트용)

OpenAI Compatible Endpoint Connection (For Agent)

Madikis

We want more video tutorials on all the features and if possible show use cases too..

Dictate Digital

Strengthen the overall stability and reliability of AgenticFlow by hardening core systems, improving error resilience, and surfacing clear diagnostics—so that every feature, from agent editing to MCP integrations, behaves predictably under load and failure conditions. Key Deliverables<ol><li>Unified Health &amp; Retry Framework<ul><li>Implement a central retry/circuit-breaker layer for all external calls (Auth0, MCP providers, webhooks) with configurable backoff and failure thresholds.</li><li>Integrate per-call timeouts and global request quotas to prevent cascading slowdowns.</li></ul></li><li>MCP Provider Fallback Registry<ul><li>Finalize the abstraction layer with automatic failover across multiple MCPs for each action.</li><li>Expose provider health metrics and allow dynamic priority reordering from the admin UI.</li></ul></li><li>Session Management Hardened<ul><li>Ensure Auth0 session renewal is bullet-proof; eliminate forced logouts during long edit sessions.</li><li>Surface session-expiry warnings to users 5 minutes before logout.</li></ul></li><li>Connector Resilience<ul><li>Apply structured JSON responses and retries for Google Sheets, Airtable, Supabase, and all direct integrations.</li><li>Add bulk-data pagination for large datasets and clear error messages on invalid inputs or permissions issues.</li></ul></li><li>End-to-End Monitoring &amp; Alerts<ul><li>Instrument key workflows with latency and error dashboards.</li><li>Configure Slack/email alerts for error rates &gt;5% over 5 minutes or average latency &gt;2 seconds.</li></ul></li><li>Documentation &amp; UX Signals<ul><li>Publish a “Stability Status” page showing real-time system health.</li><li>Add in-app banners or tooltips when a downstream service is degraded.</li></ul></li></ol><div data-type="horizontalRule"><hr></div><h3>Acceptance Criteria</h3><ul><li>All external calls use the shared retry/circuit-breaker library with sensible defaults.</li><li>MCP actions automatically fail over to a secondary provider within 500 ms of a primary failure.</li><li>No users experience unexpected logouts during 30-minute edit sessions.</li><li>Sheet/Airtable/Supabase connectors return structured JSON on the first try and retry on transient failures.</li><li>Health dashboard is live, and alerts trigger when thresholds are breached.</li><li>In-app stability notifications appear when any MCP provider or integration is degraded.</li></ul>

High Priority

Platform Stability & Reliability

Chrome extension to use the app via chrome and access it from any screen

Phần mở rộng Chrome để sử dụng ứng dụng qua Chrome và truy cập từ bất kỳ màn hình nào.

Phần mở rộng Chrome

Chrome Extension

Write to Knowledge Table Action in Workflow. Both Read From and write to long-term, we need persistent memory. Context: There is this API - docanalyzer: <a target="_blank" rel="noopener noreferrer nofollow" class="link" href="https://docanalyzer.ai/docs/api">https://docanalyzer.ai/docs/api</a> - which is very good for reviewing an extensive PDF database. One can query 100's of documents simultaneously. I was wondering if it is possible to create a workflow in order to have the AI agent store the results of the questions in its knowledge base. The idea would be that for each query, the agent builds on the existing knowledge base, so it becomes a specialist in a subject as it stores the relevant data.

워크플로우의 지식 테이블에 쓰기 작업. 장기적으로 읽기 및 쓰기 모두 영구 메모리가 필요합니다. 컨텍스트: 광범위한 PDF 데이터베이스를 검토하는 데 매우 유용한 API인 docanalyzer <a target="_blank" rel="noopener noreferrer nofollow" class="link" href="https://docanalyzer.ai/docs/api">(https://docanalyzer.ai/docs/api</a> )가 있습니다. 100개의 문서를 동시에 쿼리할 수 있습니다. AI 에이전트가 질문의 결과를 지식창고에 저장하도록 워크플로우를 만들 수 있는지 궁금합니다. 각 쿼리에 대해 에이전트가 기존 지식 베이스를 기반으로 구축하므로 관련 데이터를 저장하면서 해당 주제에 대한 전문가가 되는 것이 아이디어입니다.

워크플로우에서 지식 테이블에 쓰기 작업

Write to Knowledge Table Action in Workflow

John Bubliner

While agents can reference a dedicated knowledge base (e.g., for retrieving organizational policies, product data, or domain-specific information), workflows lack this critical feature. This prevents LLM-powered workflows from accessing structured, company-specific data during automated processes.

Adding a Knowledge Base to a Workflow System

strategix

Offer multi‑agent orchestration as a standalone purchase, without raising the base Tier&nbsp;4 price.

Bieten Sie Multi-Agenten-Orchestrierung als eigenständigen Kauf an, ohne den Basispreis für Tier 4 zu erhöhen.

Multi-Agent Add-On

Ofrezca la orquestación multiagente como compra independiente, sin aumentar el precio base del nivel 4.

Complemento multiagente

Tier 4 の基本価格を上げることなく、マルチエージェント・オーケストレーションを単体で購入できるようにする。

マルチエージェント・アドオン

기본 티어 4 가격을 인상하지 않고 독립형 구매로 멀티 에이전트 오케스트레이션을 제공하세요.

멀티 에이전트 애드온

Completed

Multi‑Agent Add‑On

Get a simple <code>&lt;iframe src="…"&gt;</code> URL for embedding agents, without needing the <code>&lt;script&gt;</code> snippet. Hi Sean, Im trying to embed it into Heartbeat Chat community in a post, so it´s not in WordPress/Divi. Im following this guide: <a target="_blank" rel="noopener noreferrer nofollow" class="link" href="https://help.heartbeat.chat/hc/en-us/articles/33257844559761-How-to-Embed-HTML-Content-in-a-Thread">https://help.heartbeat.chat/hc/en-us/articles/33257844559761-How-to-Embed-HTML-Content-in-a-Thread</a>

Direct iframe embed for agents

Register HTTP endpoints in the UI so workflows can be triggered automatically by external webhooks.

Inbound Webhook Listener

What could be beneficial is creating an AI agent for Agenticflow users to use withing the workflow builder. Train a GPT on all relevant Agenticflow documents (or add more detailed info), and instruct it to help out building workflows, trouble shoot, brainstorm etc. Simply by chatting with it (like a chat bot), maybe allow uploading screenshots or even automatic screen recording within the workflow, while the AI assistant can guide you through everything.

Lo que podría ser beneficioso es crear un agente de IA para que los usuarios de Agenticflow lo utilicen con el constructor de flujos de trabajo. Entrenar a un GPT en todos los documentos relevantes de Agenticflow (o añadir información más detallada), y darle instrucciones para ayudar a construir flujos de trabajo, solucionar problemas, aportar ideas, etc. Simplemente chateando con él (como un bot de chat), tal vez permitiendo subir capturas de pantalla o incluso la grabación automática de la pantalla dentro del flujo de trabajo, mientras que el asistente de IA puede guiarle a través de todo.

Copiloto agente AI interno

에이전틱플로우 사용자가 워크플로우 빌더에서 사용할 수 있는 AI 에이전트를 만드는 것도 도움이 될 수 있습니다. 모든 관련 에이전틱플로우 문서에 대해 GPT를 훈련시키고(또는 더 자세한 정보를 추가) 워크플로우 구축, 문제 해결, 브레인스토밍 등을 돕도록 지시하세요. 채팅 봇처럼 대화하기만 하면 워크플로우 내에서 스크린샷을 업로드하거나 자동 화면 녹화를 허용할 수도 있고, AI 어시스턴트가 모든 과정을 안내할 수도 있습니다.

사내 AI 상담원 부조종사

In house AI agent co-pilot

Wes Vg

You ahve 2500 MCP which is huge, can we use it as node in workflow automation?

Usted tiene 2500 MCP, que es enorme, ¿podemos utilizarlo como nodo en la automatización del flujo de trabajo?

Utilización de 2500 MCP como nodos en el flujo de trabajo

Use 2500 MCP as Nodes in Workflow

Quite some of us have STRAICO as Open router alternative. If you could integrate Straico as one of your API BYOK, it would greatly encourage many Sumolings to consider Agenticflow.

우리 중 상당수가 오픈 라우터 대안으로 Straico를 사용하고 있습니다. 스트라이코를 API BYOK 중 하나로 통합할 수 있다면 많은 스몰링이 에이전틱플로우를 고려하게 될 것입니다.

Straico API

Thomas Gebauer (Falkendrohne_de)

Current: CSV/Excel for knowledge baseExpect: PDF support

Current: 지식창고용 CSV/Excel기대하세요: PDF 지원

지식창고용 PDF 업로드 지원

Support PDF upload for knowledge base

Hi, OpenAI just released these models. Could you please support this in the OpenAI Nodes ?

Support GPT-4.1/4.1 mini/4.1 nano

Can you add a functionality to clone AI Agents once they are built? Oftentimes when you build one Agent you need a new one, but modified. You don’t want to start from scratch. Ability to clone AI Agents will solve that. btw. Just purchased Tier4 today &amp; loving it so far.

Ability to Clone AI Agents

Marcin Chirowski

I would love to see a native scheduling feature added to AgenticFlow.This would allow agents or workflows to be triggered automatically based on a predefined schedule (e.g., every hour, daily at a specific time, weekly, etc.), without the need for external tools.This capability would enable more autonomous and time-based operations, making the platform even more powerful for ongoing tasks and routine automations.

Ich würde es begrüßen, wenn AgenticFlow um eine native Zeitplanungsfunktion erweitert würde.Damit könnten Agenten oder Workflows automatisch nach einem vordefinierten Zeitplan ausgelöst werden (z.B. stündlich, täglich zu einer bestimmten Zeit, wöchentlich usw.), ohne dass externe Tools erforderlich wären.Diese Fähigkeit würde autonomere und zeitbasierte Vorgänge ermöglichen und die Plattform für laufende Aufgaben und Routineautomatisierungen noch leistungsfähiger machen.