01

A skill claim is a starting point

“I know Python” may be sincere, but it leaves many questions unanswered. Which concepts? At what level? In what setting? With how much support? A course completion, an examination and a software project each provide different evidence.

A more useful capability record does not force those signals into a single number. It explains what a person did, the conditions under which they did it and what someone else has reviewed.

02

Assessments need a defined target

An assessment is credible when its task actually fits the competency it claims to measure. A timed multiple-choice test may be useful for some knowledge, but it cannot stand in for every form of practical work. Speed may be relevant to one task and irrelevant to another.

The result should retain its context: the objective, assistance rules, method, issuer and any route for review or appeal. That context helps prevent a score from becoming an overconfident statement about a person.

03

Practical work needs provenance

A project can show initiative and application, but a portfolio screenshot alone may not explain a person’s contribution. A useful record identifies the project, their role, supporting material, dates and relevant skills. Where verification is claimed, it should identify who reviewed the evidence and by what method.

Collaborative work deserves particular care. “I was on the team” is different from “I designed the interface” or “I implemented the data model.” A fair record can recognize each contribution without diminishing the shared result.

04

Time is context, not a certificate

Meaningful practice takes time, so FLOOR9’s proposed Flight Hours can help describe a journey. But hours of instruction, guided practice and independent work should be distinguished. None alone proves competence.

The long-term aim is verified capability progression: evidence that a person has improved in a relevant ability. It is a demanding standard, and that is why the first pilot should start with a narrow, reviewable learning-to-assessment workflow.