How this works
Two engines, one data source. The generator turns a sector's competency statements into a program; the scorer turns a learner's responses to that program into levels. Both are deterministic — the same sector and length always produce the same modules and the same items, so a saved attempt keeps lining up with what was answered.
The four levels
Published descriptions from the Pathsmith™ Durable Skills Framework. The framework reference carries all ten clusters.
- EmergingJust beginning to learn or acquire the skill.
- DevelopingMoving in a positive direction in a professional career. Still learning, with room to grow around the skill.
- ProficientHas functional knowledge of the skill and can apply it properly in specific on-the-job settings.
- AdvancedApplies the skill naturally, without having to think about it, including under stressful or challenging circumstances.
Where the time goes
- Sector coreThe work itself — tasks, tools, settings, and standards of this sector's entry roles.30%
- Durable skillsThe ten domains, weighted by what this sector's occupations actually demand.33%
- AI at workThe AI use already documented in these occupations, and the judgment it requires.17%
- Work readinessSafety, compliance, workplace norms, and turning evidence into a hire.11%
- CapstoneOne artifact and one interview that carry the weight of the whole program.9%
Design patterns, and what each is for
The design follows the content. Procedural work with one right answer is taught and tested differently from judgment work with a defensible range.
Sector orientation
Read a short brief on the sector's real work, then mark what's new to you.
Used for: Setting, employers, entry roles, and the conditions the work happens under.
mcq · selfcheck
Guided practice
Watch it done, do it with support, then do it clean.
Used for: Procedural work with a correct method — documentation, setup, handoffs, tools.
mcq · sequence · constructed
Task simulation
Run a real task end to end, making the calls where the task branches.
Used for: Core job tasks where sequence, timing, and decision points matter.
sequence · triage · sjt · rubric
Case analysis
Work a real situation from evidence to recommendation, and show the reasoning.
Used for: Judgment content — root cause, tradeoffs, competing priorities, risk.
constructed · best-worst · mcq
Role-play
Play the interaction live with a partner, then swap and run it again.
Used for: Interaction-heavy skills — customer service, escalation, difficult news, handoffs.
rubric · sjt · selfcheck
Judgment lab
Work through short workplace situations and defend the call you'd make.
Used for: Durable skills where the right move depends on reading the situation.
sjt · best-worst · triage
AI-in-the-loop studio
Do a real sector task with an AI tool: task it, check it, decide what a person still owns.
Used for: The AI work already happening in this sector's occupations.
verify · sjt · constructed · rubric
Anchored reflection
Rate yourself against the level descriptions, then get rated, then reconcile the gap.
Used for: Self-management and character content, where self-report needs a check.
selfcheck · rubric · constructed
Retrieval drill
Short, repeated, spaced recall until it's automatic.
Used for: Safety rules, terminology, tolerances, codes — content that must be recalled cold.
mcq · sequence
Team sprint
A small team takes a real constraint and ships something inside the session.
Used for: Creativity, leadership, and collaboration, which only show up under a shared deadline.
rubric · artifact · selfcheck
Portfolio build
Produce one artifact an employer would recognize, against published criteria.
Used for: Capstone evidence — the thing that proves the rest of the program happened.
artifact · rubric · constructed
Interview simulation
A real interview with a real scorer, recorded, then run again after feedback.
Used for: The last mile — translating what you can do into what you can say you can do.
rubric · constructed · selfcheck
Scoring rules
Evidence is level-anchored. An item written at Proficient only tells you about Proficient. Getting an Emerging item right is not evidence that someone holds up under pressure.
Levels are attained, not averaged. A learner reaches a level by clearing 70% of that level's evidence with every level beneath it at 60% or better. A strong answer at Advanced does not paper over a failed Emerging item.
Thin evidence reads as thin. A domain with no evidence at a level cannot be credited above it. That shows as “not yet evidenced” rather than as a confident low score.
Self-ratings are reported, never scored. They appear next to the measured level. The gap between them is the coaching conversation, which is the reason to collect them at all.
Gates override the composite. Doing the work to standard, being dependable, being understood, and checking AI output before it counts are hard requirements. Failing one caps readiness at approaching, regardless of the rest.
Stage of life
Four stages: Student, Workforce Entry, Career Advancement, Career Transition. They are derived from four intake questions rather than asked directly, because the labels carry baggage — few people volunteer for a bucket named “Career Transition”, and “Student” gets chosen by anyone currently enrolled regardless of the twenty years of work behind them. The derived answer is always shown back with its reasoning and is always overridable.
Stage never touches a score. Identical evidence produces an identical level for every learner. The scorer computes levels with no access to stage, and stage is attached afterward. Building it the other way would produce an instrument that rates people on their biography.
What stage does change is the claim a level supports. Applying-level critical thinking demonstrated across a prior career is the same level as the same result from coursework — but the open question differs. For the career changer it is transfer into this sector's conditions and vocabulary, not acquisition. For the student it is whether the skill survives a job's pace, supervision, and consequences. The report states which question is live.
Stage also reweights the strands, sets a floor on module levels — a career changer sitting through Emerging-level instruction on a skill they built over fifteen years is wasted time — and adds one module written for that stage.
Calibration is disclosed. The framework was written and validated for people with roughly a year or less of job experience. Only Workforce Entry sits inside that population. Reports for the other three stages carry a note saying so.
Where the content comes from
Module purposes, learning outcomes, rubric anchors, and the creditable options in every judgment item are drawn from the leveled competency, durable-skill, and AI-at-work statements written for the occupations inside the sector. Distractors are built from the level below (partially creditable), an adjacent domain (plausible but off-target), and a curated anti-pattern library (the tempting shortcut). An item with three obviously wrong options measures reading speed, not judgment.