Foql for your agent

Put your agent on the next ad.

Give your agent access to Foql's competitive intelligence, brand context, and production workflow. Ask it to find an angle, develop the script, and bring back an original ad.

Agent connection is in development. This page describes the surface we are building and what it will do. Nothing here is connectable yet. See where it stands.

Use web app See agent access

Private beta. Competitive intelligence and script development are open; produced video is in limited release.

Same workspace. Same production allowance. Start either way.

One task, end to end

1Sourced findings. Competitor observations with source references and dates.
2A product-specific concept. A pattern connected to your product and audience.
3Script review. An original draft you can read before anything renders.
4Production authorization. The estimate first, then your go-ahead.
5Job status. A stable job reference your agent can check.
6A playable result. The finished video, plus the project to open in Foql.

Illustrative flow. The steps a briefed agent works through.

Produced in Foql. Pulse health tracker app.

Give it the whole brief.

Ask for the work you need done. Foql gives your agent the evidence and production capabilities to carry the idea through.

Research the market.

Find relevant competitor and creator activity with sources and dates.

Find your angle.

Connect a useful pattern to your product, including ideas from supported neighboring categories.

Make it yours.

Develop an original script using saved brand information and authorized assets.

Produce the ad.

Run the approved job and retrieve the completed video.

Bring it back.

Return the preview, export links, and project so the team can review the work.

Agent brief

“Review our tracked competitors, suggest three relevant creative angles, and draft an original script for the strongest fit. Include the sources and your reasoning. Show me the production estimate before rendering.”

See agent access →

Your agent gets the brand brief, too.

Product facts, assets, voice, and approved claims live in your Foql workspace. Your agent can use the context you've authorized, so the work starts with your brand already in the picture.

Select the right brand and product for the task.
Carry approved context from research into production.
Keep the draft, source references, and output in the same project.

Start in chat. Pick it up in Foql.

Open the same project in the web app to compare ideas, refine the script, and review the finished video. Send your agent back to the next task without rebuilding the brief.

The same project, either way
In the conversation	In the web app
Your agent returns the concept, sources, and draft script.	The same project opens with its evidence and revisions.
Authorize production and get a job reference back.	Watch the job and review the finished video.
Ask for the next task against the same brand.	Edit the script and keep the revision history.

See how the two routes meet

Set the brief. Set the limits.

See the production estimate and authorize the work you want. Your agent runs within the access and usage you give it, with status updates and reviewable outputs.

Separate research and drafts from paid production.
Authorize a single job or a bounded batch.
Review completed creative before it enters a campaign.
Manage the connection in your workspace settings.

Scoped access, usage authorization, connection revocation, and job accounting are requirements we are building the agent surface against, not features you can connect to today.

Meet your agent where it works.

Choose a supported connection and sign in to Foql. Your agent can then work with the brands and projects you've authorized.

Connection through the Model Context Protocol is the approach we are building first. We will list a client here once we have tested its setup end to end, with the steps that actually work for it. Until then the connection guide tracks the state of each one.

See connection options

Agent questions.

Is this a chatbot inside Foql?

This flow connects an agent you use to Foql’s capabilities. The web app remains a complete way to explore, direct, and review the same work.

Can I get a finished video through the agent?

Yes. Once an authorized production job completes, your agent can retrieve its preview and export links. The project is also available in Foql.

Can the agent launch my paid ads?

Foql’s launch workflow produces assets. Campaign publishing and media spend stay in your existing advertising tools.

Can it run recurring tasks?

That depends on the scheduling capabilities of your connected agent or orchestration tool. Recurring requests still use Foql’s permissions and plan limits. Built-in autonomous campaign recommendations are a separate roadmap capability.

Do I need to write code?

A supported agent connection can let you work in natural language. Setup requirements vary by client; the connection guide explains the steps.

When can I connect an agent?

The agent surface is in development and not open yet. The connection guide lists what is built and what is still to come, and it is the page that will carry the setup steps first.

Your agent has a new assignment.

Find an angle. Make it yours. Bring back the ad.

Use web app See agent access

Private beta. Competitive intelligence and script development are open; produced video is in limited release.