How the group is doing, in aggregate, against the CCAR-F blueprint. There is no per-person row on this page and there will not be one — see the rule below, which the page enforces rather than promises.
Aggregate to the cohort. Individual scores to that person and the programme lead only.
localStorage. Your own detail is on
your readiness page and stays in your browser and your D1 row.
Built this way from the start on purpose: retrofitting privacy after people have been burned does not work.
No cohort data yet. Aggregate figures come from the D1 database through an endpoint that returns counts and medians and never rows; until that endpoint lands, the readiness columns below read “—”. The study-material figures further down are real — they come from the blueprint and the item bank, which are programme facts, not personal ones.
0 of 10 by September 3, 2026
“Active” needs both halves at once: a current credential — CCAR-F is valid for 12 months — and Claude used in the past 90 days. A valid credential with no recent usage does not count. 12 seats are funded, deliberately above the target, because credentials belong to individuals and leave with them.
0 0 runs recorded
Distinct people with at least one completed practice run. A count, never a leaderboard: who is practising most is not information this page publishes.
0 report every CPN course complete
Self-reported, because Skilljar has no API we control. The professional-development bonus pays on separately verified completion through its own process — this site is not the system of record for anything that pays money.
Median percent-correct across the people who have answered items in each domain, next to what the blueprint says the domain is worth. Weight is the reason to care: a soft D1 (27%) costs roughly 1.8× a soft D5 (15%).
| Domain | Weight | Items on a real sitting | People | Median % correct | Statements at floor | Items authored |
|---|---|---|---|---|---|---|
| D1 · Agentic Architecture & Orchestration | 27% | 16.2 | 0 | — | 7 / 7 | 26 |
| D2 · Tool Design & MCP Integration | 18% | 10.8 | 0 | — | 5 / 5 | 17 |
| D3 · Claude Code Configuration & Workflows | 20% | 12 | 0 | — | 6 / 6 | 26 |
| D4 · Prompt Engineering & Structured Output | 20% | 12 | 0 | — | 6 / 6 | 24 |
| D5 · Context Management & Reliability | 15% | 9 | 0 | — | 6 / 6 | 27 |
“Items on a real sitting” is domain weight × 60 items, derived from the blueprint. “Statements at floor” counts published task statements with at least 3 practice items written against them.
Shared study material is genuinely aggregate — it belongs to the group, so it can be shown in full. This is the honest measure of collective progress while the readiness columns are still empty.
120 of 90 floor
3 items per published statement × 30 statements. Every item cites the statement it was written against.
30 of 30
The thin ones are where the cohort is collectively unprepared, regardless of anybody's score.
6 of 6 able to fill a block
A lab is ready once it can fill a 15-item block, the density a real sitting uses. A sitting presents 4 of the 6 published scenarios.
| Scenario | Domains | Weight spanned | Items authored | Block |
|---|---|---|---|---|
| S4 · Developer Productivity with Claude | D1 D2 D3 | 65% | 20 | ready |
| S1 · Customer Support Resolution Agent | D1 D2 D5 | 60% | 20 | ready |
| S3 · Multi-Agent Research System | D1 D2 D5 | 60% | 20 | ready |
| S5 · Claude Code for Continuous Integration | D3 D4 | 40% | 20 | ready |
| S2 · Code Generation with Claude Code | D3 D5 | 35% | 20 | ready |
| S6 · Structured Data Extraction | D4 D5 | 35% | 20 | ready |
They are on your readiness page, which only you and the programme lead can see. If you want a second pair of eyes on a weak domain, share it deliberately — the tool will not do it for you, and the cohort view will never surface it on your behalf.
What the cohort should share is score reports at the domain level after a real
sitting: percent-correct by domain, nothing more specific. That is what shows which study
material is thin. Item content, topics, and “there was a lot about X” are off-limits — see
INTEGRITY.md.