Builds questions and assessments that measure what they claim — matching item type to the thing being assessed, writing distractors that diagnose, and telling a question that measures learning from one that measures reading speed or test-taking. Use this to write or review assessment items, design a unit check or diagnostic, work out why scores do not match classroom performance, or decide how much evidence a claim about mastery actually needs.
The English summary is being prepared. Showing the original description.
See how to installReview the instructions the agent will receive before you install.
Every question measures something. The work is making sure it measures the thing you meant, because the alternatives — reading speed, familiarity with the format, willingness to guess — are always available and are usually easier for the student to use.
An assessment is an argument: the student did this, therefore they can do that. The argument is where assessments fail, and it fails silently.
Write down the claim first — "can decompose a two-digit number into tens and ones" — then ask what performance would be evidence for it, and what performance would be evidence against. An item that a student who lacks the skill can still get right is not evidence. An item that a student who has the skill can still get wrong, for reasons unrelated to it, is worse: it produces a false negative that gets acted on.
The most common unrelated reason is reading. Any item whose stem is harder to read than the skill is to perform has quietly become a reading assessment.
Mismatch is the usual failure: a standard describing explanation assessed by a multiple-choice item, which measures whether the student can pick an explanation someone else wrote.
In a well-built multiple-choice item, each wrong answer is the result of a specific, predictable error. Then the pattern of wrong answers says what to reteach, and the item earns its place.
Distractors that are merely wrong — a random number, an obviously absurd option — turn a four-option item into a two-option one and tell you nothing beyond right or wrong.
The mechanical tells of a weak item, all of which students learn to exploit long before they learn the content:
A single item carries noise — a misread word, a slip, a lucky guess — that swamps the signal for any individual student. Inferring mastery from one response is the most common measurement error in classroom material, and the most consequential, because it gets recorded.
Several items per claim, varied in surface form so that recognition of the format is not what is being measured. If a claim is worth recording against a student's name, it is worth three items.
Spacing matters as much as quantity: performance on a skill immediately after it is taught measures something closer to short-term recall than to learning. The same item a fortnight later measures more.
Formative assessment exists to change what happens next, which means it must be quick, frequent, low-stakes, and read immediately. An assessment that takes a week to score cannot be formative whatever it is called.
Summative assessment exists to record a judgment, which means it must be defensible: enough items, a rubric written in advance, and conditions that were the same for everyone.
Material sold as practice is usually formative in function and summative in appearance, which is
worth stating plainly to a buyer. education:learning-materials-design covers the practice
sequence this sits inside.
references/sources.md in this skill lists the outside authorities that settle the questions
here — what each one is authoritative for, and what you may do with it. Check them before
answering on anything they cover, and cite what you used. Most are free to read and not free
to reproduce; the use note on each is binding.
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Designs and audits who can reach what — authentication, authorization models, privileged access, service credentials, and joiner-mover-leaver process. Use this to design a permissions model, run an access review, reduce standing privilege, handle offboarding, set up SSO or MFA, manage service and machine credentials, or diagnose why permissions have sprawled.
The English summary is being prepared. Showing the original description.
Concentrates marketing and sales effort on a named set of accounts rather than on volume — qualifying whether the model fits your economics at all, building the account list and the buying group inside each, tiering effort against account value, coordinating so the account experiences one campaign rather than several, and measuring account progression instead of leads. Use this to decide whether to run an account-based program, build one, or work out why an existing one produces activity and no pipeline.
The English summary is being prepared. Showing the original description.
Gets new users from signup to first real value — signup flow, onboarding, time-to-value, and the early experience that determines whether someone becomes a user or a lapsed account. Use this to design or fix signup and onboarding, diagnose why signups do not convert to active use, reduce time-to-value, or decide what a new user must accomplish first.
The English summary is being prepared. Showing the original description.
Designs orchestrator-and-subagent hierarchies for a repository — splitting agents by exclusive write surface, pairing every producer with an independent auditor, and enforcing the split with a script that runs in CI. Use this whenever the user wants to set up, expand, audit, or fix a multi-agent or subagent structure for a codebase; asks how to divide work between agents; wants agent charters, roles, or a surface map written; or is hitting agents that collide on the same files, review their own work, or drift from their remit. Also use when sizing a roster or deciding whether a new agent is justified.
The English summary is being prepared. Showing the original description.
Governs models and AI systems in production — intended use, evaluation, monitoring, human oversight, documentation, and the decision to deploy or retire. Use this before deploying a model or AI feature, when defining evaluation criteria, when a model's behavior has drifted, when assessing AI risk or regulatory exposure, or when deciding whether an AI system is fit for a consequential decision.
The English summary is being prepared. Showing the original description.
Produces executive-level research — market sizing, competitor mapping, trend analysis, and strategic intelligence — grounded in cited sources with the confidence in each claim made explicit. Use this to analyze a market or industry, map competitors, evaluate a market-entry or build-versus-buy decision, produce a research brief, or assemble evidence for a decision. Also use when comparing options that need a structured, evidence-based verdict rather than an opinion.
The English summary is being prepared. Showing the original description.