Password protection for public Agents and Workflows
Agents and Workflows can only be shared with no one or everyone. There is no option to share with a team or specific individuals. It would be helpful to have the option to enable a password when sharing an Agent or Workflow.

Carsten 13 days ago
Password protection for public Agents and Workflows
Agents and Workflows can only be shared with no one or everyone. There is no option to share with a team or specific individuals. It would be helpful to have the option to enable a password when sharing an Agent or Workflow.

Carsten 13 days ago
A must add feature
In workflow there is no delay option in the workflow. Like waiting for 30 second before reaching to next step. Thank you

Harshit Garg About 2 months ago
A must add feature
In workflow there is no delay option in the workflow. Like waiting for 30 second before reaching to next step. Thank you

Harshit Garg About 2 months ago
Structuring agents and workflows through tags
It is very difficult to keep an overview, with agents and workflows everywhere. They are all in one folder, and it is hard to find connections. Which agents are active, which are in development, and which are outdated or just tests. Which workflows are part of which agents, and so on. That is even worse if you work in teams. Tags (word and color) would help much. (This would also help with knowledge files).

Carsten 2 months ago
Structuring agents and workflows through tags
It is very difficult to keep an overview, with agents and workflows everywhere. They are all in one folder, and it is hard to find connections. Which agents are active, which are in development, and which are outdated or just tests. Which workflows are part of which agents, and so on. That is even worse if you work in teams. Tags (word and color) would help much. (This would also help with knowledge files).

Carsten 2 months ago
Open multiple Claw windo for parallel processing
Hi, I would like to be able to open multiple claw windows (instances of claw) - right click and open a new window as it is possible with Antigravity, etc. So I can have 4 claw on my screen simultaneously (4 chats).

John Bubliner 2 months ago
Open multiple Claw windo for parallel processing
Hi, I would like to be able to open multiple claw windows (instances of claw) - right click and open a new window as it is possible with Antigravity, etc. So I can have 4 claw on my screen simultaneously (4 chats).

John Bubliner 2 months ago
Can CLAW freely switch between light and dark modes?
Can CLAW freely switch between light and dark modes? But I don’t want to change the default windows settings. Thanks!

DingChing Hong 2 months ago
Can CLAW freely switch between light and dark modes?
Can CLAW freely switch between light and dark modes? But I don’t want to change the default windows settings. Thanks!

DingChing Hong 2 months ago
Support Custom Model IDs for OpenAI-Compatible Connections
Currently, Agentic Flow AI's OpenAI connection only allows selecting models from a pre-defined list. This makes it impossible to use custom OpenAI-compatible APIs (like Alter) that require specific model IDs (e.g., anthropic#claude-3-5-sonnet-20241022). To enable wider compatibility and flexibility, please add the ability to: a) Enter a custom model ID in the OpenAI connection settings, OR b) Add a "Custom Model" field in the agent configuration. This would allow users to connect to a broader range of LLMs and leverage services like Alter, which provide access to multiple frontier models through a single API key."

bourbonblack 3 months ago
Support Custom Model IDs for OpenAI-Compatible Connections
Currently, Agentic Flow AI's OpenAI connection only allows selecting models from a pre-defined list. This makes it impossible to use custom OpenAI-compatible APIs (like Alter) that require specific model IDs (e.g., anthropic#claude-3-5-sonnet-20241022). To enable wider compatibility and flexibility, please add the ability to: a) Enter a custom model ID in the OpenAI connection settings, OR b) Add a "Custom Model" field in the agent configuration. This would allow users to connect to a broader range of LLMs and leverage services like Alter, which provide access to multiple frontier models through a single API key."

bourbonblack 3 months ago
Accounting MAS
Hi I was wondeirng if there could be a template or more documentaiton on how to set up our own MAS team. I am very interested in creating my own MAS Accounting team for myself. To keep track of my expenses etc. I think if I can receive some support on how to get each individual agent up and running it will be great. Appreciate you guys!

NinjaPumpkin 3 months ago
Accounting MAS
Hi I was wondeirng if there could be a template or more documentaiton on how to set up our own MAS team. I am very interested in creating my own MAS Accounting team for myself. To keep track of my expenses etc. I think if I can receive some support on how to get each individual agent up and running it will be great. Appreciate you guys!

NinjaPumpkin 3 months ago
UI Update - Better Workflow Management
It’d be great if there was an update to the way we can see the workflows on the top level of the account. As it stands i can’t tell which workflows are running or not, there’s no easy way to toggle things on or off. It says when it was last made or modified but not when it was last run, or if a run has encountered an error. I think having a bit more insight into the workflows from the top level of the manager screen would be a huge ux improvement for many reasons. It’d be great if the workflow and agents could be a table or little card that showed Enabled/Disabled, Last Run, Status ie (Ran without errors or ran with errors), Monthly token cost and maybe some options to rename instead of duplicate or deleate being the only options. Lot of functionality missing here that should be easily accessable from the top level.

Taylor Thain 3 months ago
UI Update - Better Workflow Management
It’d be great if there was an update to the way we can see the workflows on the top level of the account. As it stands i can’t tell which workflows are running or not, there’s no easy way to toggle things on or off. It says when it was last made or modified but not when it was last run, or if a run has encountered an error. I think having a bit more insight into the workflows from the top level of the manager screen would be a huge ux improvement for many reasons. It’d be great if the workflow and agents could be a table or little card that showed Enabled/Disabled, Last Run, Status ie (Ran without errors or ran with errors), Monthly token cost and maybe some options to rename instead of duplicate or deleate being the only options. Lot of functionality missing here that should be easily accessable from the top level.

Taylor Thain 3 months ago
I want to give a full 5-star rating to the AgenticFlow AI team for their outstanding support. Their responsiveness, expertise, and helpfulness made a big difference in my experience. Thank you for the excellent service!

PlaystationDZ 3 months ago
I want to give a full 5-star rating to the AgenticFlow AI team for their outstanding support. Their responsiveness, expertise, and helpfulness made a big difference in my experience. Thank you for the excellent service!

PlaystationDZ 3 months ago
Feature Request: Increase the Recursion Limit upper #
The Recursion Limit is an important feature, but I am finding that it is inhibiting my agents… 100 is too low of a max I am finding at times…

M Scott 3 months ago
Feature Request: Increase the Recursion Limit upper #
The Recursion Limit is an important feature, but I am finding that it is inhibiting my agents… 100 is too low of a max I am finding at times…

M Scott 3 months ago
Please add the ability to edit the workspace name.
Please add the ability to edit the workspace name. Currently, it is automatically generated during registration (in my case, taken from my First and Last Name in Gmail). I would like the option to specify my own brand or domain name.

Pavel Tolochko 3 months ago
Please add the ability to edit the workspace name.
Please add the ability to edit the workspace name. Currently, it is automatically generated during registration (in my case, taken from my First and Last Name in Gmail). I would like the option to specify my own brand or domain name.

Pavel Tolochko 3 months ago
Get the LLM action/node updated with Gemini Flash 2.5 and GLM 4.x
Right now these models are not exposed in this action. Since Sean has said he wants to really dial into GLM Air 4.5 as the “daily driver”, it would be nice to have this action and whatever other ones that use/rely on Agentic Flow native models to get the more reliable models attached… Screen shot shows what is in the action model list currently…

M Scott 3 months ago
Get the LLM action/node updated with Gemini Flash 2.5 and GLM 4.x
Right now these models are not exposed in this action. Since Sean has said he wants to really dial into GLM Air 4.5 as the “daily driver”, it would be nice to have this action and whatever other ones that use/rely on Agentic Flow native models to get the more reliable models attached… Screen shot shows what is in the action model list currently…

M Scott 3 months ago
Allow users to use Built-in Plugins
The built in plugins a don’t work, user need to select PixelML plugins costing them paid credits.

Mario Ortiz 3 months ago
Allow users to use Built-in Plugins
The built in plugins a don’t work, user need to select PixelML plugins costing them paid credits.

Mario Ortiz 3 months ago
Monitoring Page
Have a monitor/reporting area to be able to see the credits consumed on usage, with enough details to be able to track it down (Agent, Workflow, Model, etc.), similar to the Usage History panel inside the Settings area. As discussed in the Office Hours #29

lesthertod 4 months ago
Monitoring Page
Have a monitor/reporting area to be able to see the credits consumed on usage, with enough details to be able to track it down (Agent, Workflow, Model, etc.), similar to the Usage History panel inside the Settings area. As discussed in the Office Hours #29

lesthertod 4 months ago
don't use credits if the action fail
If I try to generate an image and it fail or something else fail, the credits should not be used.

Matteo Raggi 4 months ago
don't use credits if the action fail
If I try to generate an image and it fail or something else fail, the credits should not be used.

Matteo Raggi 4 months ago
WebSocket (WSS) Endpoint Support for Phone Agents
Hi, I’d like to integrate my AI Agents with telephony platforms such as Retellai.com. To make this possible, the agent needs to be exposed as a WebSocket Endpoint (WSS) instead of using Webhook (API POST). I kindly ask you to consider implementing this functionality. Thank you.

Devid 5 months ago
WebSocket (WSS) Endpoint Support for Phone Agents
Hi, I’d like to integrate my AI Agents with telephony platforms such as Retellai.com. To make this possible, the agent needs to be exposed as a WebSocket Endpoint (WSS) instead of using Webhook (API POST). I kindly ask you to consider implementing this functionality. Thank you.

Devid 5 months ago
How much knowledge base in use?
add a way to see how much of the space u use. I have 2gb , but how do i know how much is left?

Henk Overdijk 5 months ago
How much knowledge base in use?
add a way to see how much of the space u use. I have 2gb , but how do i know how much is left?

Henk Overdijk 5 months ago
Completed
add more then 10 knowledge files to agend
Right know we can only have 10 knowledge files per agend. I would love to be able to add a whole lot more. This way i can make it into a tool for researchers or students. It would also help me justify having more knowledge space.

Henk Overdijk 5 months ago
Completed
add more then 10 knowledge files to agend
Right know we can only have 10 knowledge files per agend. I would love to be able to add a whole lot more. This way i can make it into a tool for researchers or students. It would also help me justify having more knowledge space.

Henk Overdijk 5 months ago