Your agent defines the question
It prepares a focused study you can review before real users choose to open it inside your product.
User feedback for AI agents
Your agent defines what it needs to learn. UserTold interviews real users inside your product and returns source-linked evidence your product or coding agent can act on.
Interview evidence
What the user said
"I tried this flow three times and still cannot find where to change billing settings."
What returns to the agent
Billing settings are hard to find from the current flow.
The missing feedback loop
UserTold gives product agents a direct, reviewable connection to the people using what they build.
It prepares a focused study you can review before real users choose to open it inside your product.
UserTold captures voice, screen, actions, page context, and answers from the real product experience.
Quotes and observed behavior stay linked to their source, ready for a human or project-aware agent to verify and act on.
Evidence, not a generated opinion
Every finding stays connected to what a real user said or did, so the interpretation can be checked before it becomes work.
Source moment
“I expected billing to live under account settings, not workspace settings.”
Interview replay, transcript context, and the exact page path stay attached so a reviewer can validate the interpretation.
Returned to the agent
The quote, page path, and observed behavior remain attached. A human or project-aware agent verifies the evidence before routing work.
MCP · CLI · REST API · GitHub · Linear
See how agents use UserTold →Give your agent a way to ask
Run an in-product interview, review the source-linked evidence, and give your agent a grounded next step.