All Learning

Score AI fluency across six skills — foundations, verification, prompting, workflow, safety, and role use — from your own 0–5 marks. The weakest skill is the headline (MIN, not the flattering average), and an essential gate holds anyone back to AWARE who can't verify a fact or handle data safely (Verification or Safety < 3/5) no matter how strong the rest. Tiers FLUENT / CAPABLE / AWARE, an overconfidence flag where self-rating outruns measured skill, and a per-person route to the exact training next — plus a team roll-up (weakest-learner tier + most-common gap). Runnable Python engine + a workbook that reproduces it + individual & team playbooks. For individual development and training planning — not a hiring, promotion, performance-review, or certification tool, and not legal advice.

Prove your AI-literacy program is defensible — role by role — before you spend on more training. Read your own tool inventory, role-to-system mapping, and training records; it scores coverage and proportionality per role (EVIDENCED / PARTIAL / UNEVIDENCED) on weighted levers, with a hard gate that flags any in-scope role whose staff operate an AI system with zero training record. The org verdict is the weakest in-scope role (MIN-as-headline — DEFENSIBLE / SHORTFALL / NOT DEFENSIBLE), not a coverage average that masks the one untrained operating role; coverage % and mean are context only. Date-agnostic (it scores evidence, never a statutory deadline). Runnable Python engine + a workbook that reproduces it + two playbooks (evidence record + remediation runbook). The evidence gate that runs before the AI Literacy & Workforce Training Kit — prove it, then train. Produces the audit-ready evidence record; does not certify compliance and is not legal advice.

Practice prompts like a skill — graded drills with spaced repetition. Each prompt you write is scored 0–5 on the six RCTFCV criteria (Role, Context, Task, Format, Constraints, Verification-ask) on weights summing to 100, banded STRONG / WORKABLE / WEAK — with one hard gate: if the Verification-ask is below floor (you never ask the model to check its own work), the drill caps at WORKABLE no matter how high the total. A great prompt that doesn't ask for verification can't be STRONG. Then a Leitner spaced-repetition schedule (boxes 1–5 → review in 1/2/4/8/16 sessions, no calendar dates) puts the weak drills back in front of you until they stick. Grades prompt craft, not the model's output; deterministic and offline. Runnable Python engine + a workbook that reproduces it + two playbooks (the RCTFCV rubric + a deliberate-practice guide). The skill-building companion to the AI Fluency Diagnostic. Not legal, medical, or professional advice.

Test whether you catch errors in AI output instead of trusting it. Review a piece of AI text deliberately seeded with six planted problems — a confidently-stated fabricated statistic, plausible-but-false claims, and biased framing — and mark which you caught; it scores your weighted catch-rate and returns INDEPENDENT / ASSISTED / RELIANT, with a hard gate on the fabricated statistic (miss the fake number and you can't be INDEPENDENT, capped to ASSISTED) and an over-confidence flag where you were highly confident on an item you missed — the over-reliance signal. A team roll-up reports the weakest-learner tier and the most-common missed error type. Deterministic and offline — your own marks drive the score, no AI grades you. Runnable Python engine + a workbook that reproduces it. The deep dive on the verification skill the AI Fluency Diagnostic flags. For individual development and training only — not a hiring, promotion, performance-review, or ranking tool, and not legal advice.

Backward-design an AI-adoption curriculum and grade whether it holds together. Map each target outcome to a role, a sequence step, an hours cost, and a Kirkpatrick measure, set a weekly time budget and rollout window, and the engine scores four weighted dimensions (role fit 25 · sequence & spacing 25 · outcome-measure coverage 30 · time-budget realism 20) into a 0–100 coherence score — COHERENT / LOOSE / FLAWED — and names the dimension to fix first. A hard unmeasured-outcome gate fails any plan with a single outcome that has no measure, regardless of the total (the shipped sample scores 94 and is still FLAWED). Runnable Python engine + a workbook that reproduces it + two playbooks (backward design, rollout sequencing) + a worked sample. It grades the structure of a plan you enter — it does not score or rank people and is not for any employment decision. Educational; not legal, HR, or financial advice.

Run six AI safe-use drills past your team and find out if they catch the everyday mistakes — pasting client PII into a public tool, leaking a secret or credential, shadow AI, a hallucinated fact, prompt injection from pasted content, an over-permissioned connector. Mark recognition and correct action 0–5 per drill; the weaker mark caps the drill (you can't respond to a threat you never spotted) into a SAFE / RISKY / UNSAFE verdict, and the team band is your single weakest drill — DRILLED / UNEVEN / RAW — never the flattering average. A hard regulated-data gate forces the whole team to RAW if any regulated drill lands UNSAFE (the shipped sample averages 73 and is still RAW on one missed drill). Runnable Python scorer + a workbook that reproduces it + two playbooks (running the drills, the six failure modes) + a worked sample. It scores the drills, not people — not a score of individuals, not monitoring, not a security audit. Educational; not legal advice.

A 60-minute leadership AI-readiness diagnostic. Your leadership team scores its own organization across five weighted dimensions — strategy & business case (20), spend & ROI visibility (20), risk & regulatory exposure (25), team capability & adoption (20), competitive & market awareness (15) — for a weighted 0–100 total banded LEADING (≥75) / CATCHING UP (50–74) / BLINDSIDED (<50), and the dimension to fix first. A hard critical-blindspot gate forces BLINDSIDED if either gate dimension (risk/regulatory or competitive awareness) is scored 0, no matter how high the total — the shipped sample scores 53 and is still BLINDSIDED on a single zero. Runnable Python engine + a workbook that reproduces it + a facilitator playbook (the 60-minute briefing) + a five-lesson decision course + a worked read. It grades a leadership team's preparedness, not any individual, and not for any employment decision; not legal, financial, investment, or compliance advice — confirm exposure with counsel.

A tracker for peer-led AI skill-sharing that tells you, honestly, whether the habit is taking root. Score each teach-back session on three dimensions — accuracy, transfer (peers can now do it unaided), and participation — capped by the weakest (MIN) into SOLID / PATCHY / HOLLOW, then roll the program up by density and recency to SELF-SUSTAINING / PROPPED-UP / DORMANT. A keystone gate forces DORMANT when the program is too new or the last two sessions both went HOLLOW (the shipped sample reads 0.6 solid-share but DORMANT — it has gone quiet), and an independence downgrade drops a one-teacher program to PROPPED-UP. Runnable Python engine + a workbook that reproduces it + a champion playbook + a repeatable 45-minute session kit + a worked program log. It grades a session's quality and a program's habit, not any individual teacher or learner, and not for any employment decision; not legal, HR, or compliance advice.

A one-afternoon, role-based AI micro-brief for sales reps — five quizzed lessons scored CLEARED / REVISIT / REPEAT, with an advance gate where the weakest module gates the whole brief. Three ship-this-week workflows. One .xlsx.

A one-afternoon, role-based AI micro-brief for recruiters — five quizzed lessons (bias-checked JDs, honest outreach, lawful questions, interview prep, candidate comms) scored CLEARED / REVISIT / REPEAT, with an advance gate. Grades the recruiter's skill, never a candidate. One .xlsx.

A one-afternoon, role-based AI micro-brief for bookkeepers — five quizzed lessons (categorizing, reconciliation prep, receipt capture, month-end flags, client notes) scored CLEARED / REVISIT / REPEAT, with an advance gate. AI drafts; you keep every number. One .xlsx.

A one-afternoon, role-based AI micro-brief for founders — five quizzed lessons (triage, voice-matched replies, delegation, scheduling, weekly roundup) scored CLEARED / REVISIT / REPEAT, with an advance gate. AI drafts; it never sends. One .xlsx.

A one-afternoon, role-based AI micro-brief for support reps — five quizzed lessons (ticket triage, on-brand replies, the honest escalation rule, KB & macro building, reply QA) scored CLEARED / REVISIT / REPEAT, with an advance gate. AI drafts; a person is always available. One .xlsx.

A one-afternoon, role-based AI micro-brief for executive assistants — five quizzed lessons (scheduling & calendar defense, travel, briefing prep, the exec's inbox, meeting prep & follow-up) scored CLEARED / REVISIT / REPEAT, with an advance gate. AI drafts; you confirm and send. One .xlsx.

A one-afternoon, role-based AI micro-brief for marketers — five quizzed lessons (campaign angles, on-brand copy, repurposing, channel-fit, honest measurement) scored CLEARED / REVISIT / REPEAT, with an advance gate. AI drafts; you verify every claim and disclose #ad. One .xlsx.

A one-afternoon, role-based AI micro-brief for agents — five quizzed lessons (Fair-Housing-checked listings, CMA narrative, marketing kit, lead follow-up, client comms) scored CLEARED / REVISIT / REPEAT, with an advance gate. Grades the agent, never a buyer; never invents a comp. One .xlsx.

A one-afternoon, role-based AI micro-brief for project managers — five quizzed lessons (status updates, meeting notes → action items, scope & change requests, risk & dependency logging, stakeholder comms) scored CLEARED / REVISIT / REPEAT, with an advance gate where the weakest module gates the whole brief. Plus a team cohort dashboard — one REPEAT flips the team to RETRAIN even at a high mean. One .xlsx.

A one-afternoon, role-based AI micro-brief for office managers & admins — five quizzed lessons (internal announcements, vendor & facilities coordination, meeting & event logistics, expense & document organization, onboarding & handbooks) scored CLEARED / REVISIT / REPEAT, with an advance gate where the weakest module gates the whole brief. Plus a team cohort dashboard — one REPEAT flips the team to RETRAIN even at a high mean. One .xlsx.

A one-afternoon, role-based AI micro-brief for operations managers — five quizzed lessons (SOPs that get followed, process docs & handoffs, vendor & supplier comms, scheduling & shift notes, incident & exception write-ups) scored CLEARED / REVISIT / REPEAT, with an advance gate where the weakest module gates the whole brief. Plus a team cohort dashboard — one REPEAT flips the team to RETRAIN even at a high mean. One .xlsx.

A one-afternoon, role-based AI micro-brief for financial advisors — five quizzed lessons (client meeting prep, plain-language explainers, review-letter & summary drafting, prospecting & client comms, verify & route for compliance) scored CLEARED / REVISIT / REPEAT, with an advance gate where the weakest module gates the whole brief. Recommends nothing and scores no clients; every draft routes through firm review (FINRA/SEC scope note). One .xlsx.

A one-afternoon, role-based AI micro-brief for freelancers & solopreneurs — five quizzed lessons (proposals that win, scope docs that prevent creep, client emails that keep trust, invoices & the polite chase, case studies & portfolio) scored CLEARED / REVISIT / REPEAT, with an advance gate where the weakest module gates the whole brief. Plus a team cohort dashboard — one REPEAT flips the team to RETRAIN even at a high mean. One .xlsx.

A one-afternoon, role-based AI micro-brief for HR generalists supporting existing staff — five quizzed lessons (policy plain-language rewrites, onboarding & handbook drafting, employee communications, PTO & benefits FAQ drafting, verify-before-you-publish) scored CLEARED / REVISIT / REPEAT, with an advance gate where the weakest module gates the whole brief. Scores skills, not employees; not for any employment decision (EEOC scope note). One .xlsx.

A one-afternoon, role-based AI micro-brief for nonprofit coordinators — five quizzed lessons (grant narrative drafting, donor thank-yous & appeals, volunteer & community comms, event & campaign copy, verify-before-you-send) scored CLEARED / REVISIT / REPEAT, with an advance gate where the weakest module gates the whole brief. Scores skills, not people; no donor data. One .xlsx.

A one-afternoon, role-based AI micro-brief for teachers & educators — five quizzed lessons (lesson outlines & objectives, rubric & assessment drafting, feedback & comment drafting, parent & guardian communication, verify-before-you-teach) scored CLEARED / REVISIT / REPEAT, with an advance gate where the weakest module gates the whole brief. Scores skills, not people; no student data. One .xlsx.

Every check your app makes, read off in three no-judgement observations — what it guards, whose computer it runs on, what it decides from — routing each to ENFORCED / TRUSTS THE BROWSER / FINE IN THE BROWSER. A server-side check that reads a browser-sent value still reads exposed. Rolls up to LINE HOLDS / LINE LEAKS / NO LINE and names the row to fix first. One .xlsx; grades the build, never a person.

Row-level security read off one row per table and per operation — because “RLS enabled” is a badge, not an answer. A policy that admits everyone still reads OPEN TABLE; a browser-reachable service-role key rewrites the whole matrix to PROJECT OPEN. Rolls up to TABLES LOCKED / GAPS IN THE POLICIES / PROJECT OPEN and names the row to fix first. One .xlsx; grades the config, never a person.

A login says who you are, never whether this record is yours — two different questions. One row per route, four observations, routing to AUTHORIZED / MEMBERSHIP IS ENOUGH / LOGGED-IN ONLY / NOBODY IS CHECKING. A signed-in-only route taking the record id from the request drops to the floor. Rolls up to PERMISSIONS HOLD / SIGNED IN IS DOING TOO MUCH / ANYONE CAN WALK IN. One .xlsx; grades the build, never a person.

Your checkout page working is not your money path working. One row per step — does money change hands, does the processor’s webhook trigger it or the browser landing on a page, where’s the amount from, would running it twice matter. Routes CONFIRMED / DISPLAY ONLY / TAKEN ON TRUST / NOT HANDLED. Rolls up to MONEY PATH HOLDS / PATH CAN BE FORGED / PATH BREAKS. One .xlsx; grades the flow, never a person.

Some keys are meant to ship in the page; some must never leave your server, and nothing about the string tells you which. One row per key: which kind, where it lives, whether it’s ever been reachable, whether it’s been rotated. Relocation is not remediation — a once-exposed secret reads ROTATE NOW wherever it now sits. Rolls up to KEYS HELD / KEYS MISPLACED / KEYS BURNED. One .xlsx; never asks for a key value.

Almost everything an app plugs into has a sandbox and a real one, and nobody wires them all in one sitting — so they drift. One row per service: which mode your work reaches, which your live app reaches, whether they share an account. Routes SEPARATED / NO SEPARATE MODES / PROD ON TEST / DEV ON LIVE. Rolls up to LINES SEPARATE / POINTING AT THE SANDBOX / WIRES CROSSED. One .xlsx; grades the wiring, never a person.

Every table holds rules — rows belong to someone, an email is unique, a price is money — and the only question is who enforces them, the table or you. One row per table read straight off your table editor. A records table with no owner column reads NOTHING TO ENFORCE AGAINST however clean the rest. Rolls up to MODEL HOLDS / YOU ARE THE CONSTRAINT / NOTHING TO BUILD ON. One .xlsx; sees no rows, grades the shape.

For the moment an AI tool hands you a change and the only button is Accept. It never judges the code — it asks the question you can answer: if this is wrong, can you get back? One row per pending change; on a schema or money change a restorable copy is required. Routes SAFE TO ACCEPT / NOTHING AT RISK / CHECKPOINT FIRST / DO NOT ACCEPT YET, rolling up to YOU CAN GET BACK / SAVE BEFORE YOU ACCEPT / ONE-WAY DOOR. One .xlsx.

The day you added a login you started holding things that belong to other people. This is an inventory of what you ended up with — it names no law and makes no legal determination. One row per field: is it a person’s information, does a feature use it, do uncontrolled copies exist, could you delete it. Deleting the row isn’t getting rid of the value if it’s also in a log. Rolls up to YOU CAN ACCOUNT FOR IT / MORE THAN YOU NEED TO HOLD / HOLDING IT FOREVER. One .xlsx; not legal advice.

Nine areas of a small app, two questions each: can you name it, and have you seen it yourself. It ranks GOING ON BELIEF below CANNOT EXPLAIN IT — a gap you admit is already on your list; a belief you never checked is the one you skip. Routes YOU CAN EXPLAIN IT / DOES NOT APPLY / CANNOT EXPLAIN IT / GOING ON BELIEF, rolling up to YOU KNOW YOUR OWN APP / YOU KNOW WHAT YOU DO NOT KNOW / GOING ON BELIEF. The line’s map — names no product. One .xlsx.

Keep the AI skills you worked to build. Each skill carries a Leitner box and a last-reviewed date; the tracker returns FRESH / FADING / COLD from your own dates against the skill's expanding interval, with a critical-cold gate that escalates the week when an essential skill lapses. Fixed eval date, no AI. One .xlsx, deterministic.

Turnkey, plain-language AI literacy program — 28-slide deck, workbook, knowledge check, certificate, and a completion log. Role-based, deploy in a week, built to evidence EU AI Act Article 4 measures. Educational, not legal advice.

A short, reassuring video course — plain English, no rush, a real safety module, and a printable companion guide. Yours forever, no subscription.

Ten drop-in Claude Project blueprints — one per business function. Sales, founder, support, marketing, ops, finance, HR, product, legal, content. Custom instructions + KB map + starter prompts each.

Master prompt + 8 focused prompts + 24-question bank across Sales/Marketing/Finance/Ops. Five-rule honesty layer keeps the output defensible. For any AI tool that runs code on uploads.