AI agent for user experience designers
Prototype Usability Session Agent
Consistent, consent-respecting usability session records.
What it does
Moderated prototype tests involve repetitive setup and note-taking that pulls the researcher's attention away from the participant. When a session starts, this agent confirms what the participant consented to. If recording was not agreed, it runs the session without recording. It presents each approved task and logs only permitted interactions. When the participant stalls, it offers only the next pre-approved neutral prompt from the script, never a hint of its own. If consent for recording is withdrawn mid-session, it stops recording at once. It assembles a timeline of observations for the researcher. The researcher owns all findings and product decisions. Edge case: a participant declines recording, so the agent logs clicks only.
How it works
Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.
Read the steps as a list
- Usability session starts
- Read the participant's consent record
- Has the participant consented to recording?If not: set the session to run without recording. Back to step 2.
- Present the next task and log interactions
- Is the participant progressing?If not: offer the next approved neutral prompt or record a stall. Back to step 4.
- Stop recording at once if consent is withdrawn
- Assemble the observation timeline
- Researcher reviews the session recordThe agent waits here for your OK.
- Usability session record for the researcher
How it decides
It picks prompts from the approved script based on observed progress and stops at protocol limits.
- Use only pre-approved neutral prompts.
- Consent controls what is recorded at every moment.
- Stop at protocol limits, even if tasks remain.
Make it yours
Every agent is a starting point. You choose these settings for your own situation.
- Approved task script and neutral prompts
- Stall time before a prompt (default 60 seconds)
- What may be logged under each consent level
- Timeline format for the researcher
What keeps you in control
It always asks you first
- Protocol changes
- Recruitment
- Product decisions
Hard limits
- No leading prompts.
It stops when
- Done: tasks complete or protocol limit.
Set it up
We guide you through the set-up, step by step
Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.
- One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
- The agent then walks you through connecting your own data, one source at a time
- A downloadable copy with the flow chart, the rules and the full guide