First demonstration.
Tell us the starting condition and intended result. Perform the skill naturally and explain anything important without changing how you work.
How it works
How a skill becomes an asset
A guided path from what you already know to a structured Skill Seed that can be adapted for compatible robots.
- 01
Apply.
Tell us what you can do.
- 02
We select.
We evaluate distinctiveness, capture feasibility, safety, future robot fit, and expected demand. Not everyone gets in.
- 03
You capture.
Complete guided sessions using the phone, tablet, or computer you already own. We walk you through the process.
- 04
We build the Skill Seed.
Your video, audio, decisions, success criteria, and failure conditions become a structured record of how you work.
- 05
We test whether it can teach.
Another qualified person tries to reproduce the result from the derived specification. If important knowledge is missing, we return to the evidence and fill the gap.
- 06
We adapt and validate it.
When compatible robots are available, we work to make the skill playable and test whether a specific robot configuration can reproduce the defined result safely and reliably.
- 07
Validated skills enter the marketplace.
The skill is listed with creator attribution unless the creator chooses an approved alternative.
- 08
You share in what it earns.
You receive 50% of the net licensing revenue HumanLore receives from the skill, for the term defined in your agreement.
The capture
What a session actually looks like
You already own the equipment. One phone holds a fixed anchor view. A second phone, tablet, or laptop captures another fixed view. Built-in microphones record what you explain before, during, and after each pass.
Second demonstration.
Show the same skill again. The conditions may be similar or may include a real variation. Tell us what changed and why you adjusted.
Third demonstration.
Give us another example of the same result. Point out corrections and anything you could feel or hear that the cameras might miss.
Three demonstrations start the skill. They do not automatically finish it.
HumanLore compares the recordings, proposes what it learned, and asks you to correct its understanding. It then requests only the next question, view, measurement, boundary, failure, or recovery needed to fill an important gap.
The resulting Skill Seed holds evidence-linked video, audio, transcripts, task structure, decisions, success criteria, variation, boundaries, failures, recoveries, confidence, and unresolved gaps.
What would you teach?