Record a process. Get a Claude skill.

Show WIQ how the work gets done and we will turn your recording into a shareable Claude skill.

An example blueprint

Built from a six-minute recording of a support-ticket walkthrough - yours will look like this, for your process.

What you get with WIQ

This is a Claude skill - a solid start. A WIQ blueprint carries the same process plus the controls that make it run reliably, every time.

In this skill

Steps, inputs, decisions

Your download carries the full process - the steps, inputs, and decisions WIQ extracted from your recording - as a structured Claude skill you can hand to anyone.

In this skill

Connect the tools, or go manual

Skills execute through Claude connectors. Wire up the ones the process needs - or mark those steps for manual or in-browser execution.

WIQ only

Escalation paths, enforced

The skill above lists them; on WIQ they're checked after every step, so the agent stops and hands off the moment something looks wrong.

WIQ only

Guardrails, supervised

A downloaded skill can only ask Claude to follow the rules. WIQ enforces them - supervising the agent's reasoning at every step, not just its output.

Run your skill

Download skill file
  1. 1. Download the file.
  2. 2. Drag it into Claude and hit run - or just say “run this file.”
  3. 3. Claude sets it up, runs the process, and offers to save it as a reusable skill for next time.
  4. 4. Want to change it? The file is plain markdown - edit any step in a text editor, or tell Claude what to change and let it rewrite the file for you.
An operations lead at a desk, the WIQ app's privacy and blocklist settings open on the laptop beside her
The WIQ browser extension, showing recording active on app.getwiq.ai with its blocklist and control mode

Governance and controls for every agent you deploy

Decide what WIQ can see, what each agent is allowed to do, and where it runs. Every run is checked against the approved blueprint and logged.

Book a demo

Capture controls

Allowlist and blocklist by app, domain, and time window. Nothing outside the policy is ever recorded.

Permissions and approvals

Scope the tools each Agent can reach and the actions it can take. Set the approval gates and escalation paths before it runs.

Conformance monitoring

Every run is checked against the approved blueprint. Deviations are flagged and exceptions route to the right person.

Deployment isolation

Run locally where the work happens or inside your own cloud. The work never has to leave your environment.

SOC 2 Type II
GDPR
CCPA

Frequently asked questions

How does WIQ learn our processes?

WIQ combines observed work, existing documentation, and input from your team to build process blueprints.

What is a blueprint?

A blueprint describes the process steps, required tools, and approval points. Your team reviews it before deployment.

Where do the agents run?

Agents run in your existing AI platform using WIQ blueprints. Deployment depends on the tools and access each workflow needs.

Can we control what WIQ records?

Admins and users can define which apps and sites WIQ records using allowlists and blocklists.

What happens when our process changes?

WIQ surfaces process changes and helps your team review and approve blueprint updates before deployment.

An overhead view of a container terminal, stacks in long rows with trucks moving between them

Schedule a discovery session with an AI architect.

Book a demo