New kit is live: The Rhythm Audit.

It came out of today's study. 76,485 students, a month of log data, and a finding that AI use tracks the study schedule people already keep rather than building them a new one.

So the kit starts where the evidence starts, with the week you actually had.

Seed, 15 minutes. Mark the three blocks last week where you did focused work, and be honest about which one got eaten.

Sprout, 20 minutes. Name the one task that lives in the biggest block, what finishing it well requires of you, and which part of it is mechanical.

Grow, 30 minutes. Run that task once with AI in the loop, at the real hour, on the real work. Clock it against your normal run.

Harvest, 40 minutes. Keep, change, or drop, in one sentence. Name the block it now lives in permanently. Hand it to one colleague with the line you want challenged.

Twelve fields, a progress bar, and a copy button so you can paste the whole thing somewhere useful. Free, no gate beyond the community.

Open it: https://www.harvestkernel.com/rhythm-audit-kit/

Bring answer 12 back here and the room will argue it with you.

New kit: The Assignment Dial

Four sittings, fifteen working fields, about an hour and three quarters. You set an AI value on one real task, write the reason, then do a second task and find out what actually moved the dial between them.

Seed, 15 minutes. Name what the task proves.
Sprout, 20 minutes. Set the value and argue the opposite.
Grow, 30 minutes. Write the two sentences people will actually read on the task itself.
Harvest, 40 minutes. Do a second task, compare, hand it to a colleague.

Field 14 is the one that matters. Two tasks disagreeing about something is how your real decision rule surfaces, and most people have never written theirs down.

Nothing saves or sends. Use Copy everything at the bottom and keep it wherever you keep notes.

Open the kit: https://www.harvestkernel.com/assignment-dial-kit/

New kit: The Standing Row

A four week AI practice for working instructors. One task, one colleague, about 105 minutes total.

https://www.harvestkernel.com/standing-row-faculty-kit/

What is inside:

Seed, 15 minutes. Pick the row. Three fields that force you toward the recurring task instead of the exciting project, plus an honest timing of what it costs you now.

Sprout, 20 minutes. Run it twice on real student work. Four fields that capture what the tool got wrong, what you changed, and the one sentence that measures your own fluency.

Grow, 30 minutes. Route the hour before you have it. This is the phase everyone skips. Four fields, including the routing sentence in a fixed shape, plus the question of how a student would notice by October.

Harvest, 40 minutes. Hand it to a colleague. Four fields, and the most useful one is where you write down their objection without defending yourself.

Fifteen working fields, a progress bar, and a copy button that hands you the whole thing as plain text for your notes.

It is free, it requires no purchase, and every phase runs on any chatbot you already have. That is deliberate. If a framework only works when you buy something, it was a sales pitch wearing a lab coat.

Run it for four weeks, then come back and tell me which row you picked.

Companion kit for today's post is live. The Core Grammar Drill: 80 minutes, four SeedStacking phases, seven working fields, and a verdict log that computes your personal miss rate with AI as you go.

Seed pulls one claim you accepted this week. Sprout makes the model surface the probability under its certainty. Grow builds the case against. Harvest chases the claim to a primary source and logs the verdict. Use the export button to copy your log before you leave the page.

Run it, then post your miss rate in this thread. Honest numbers only, the embarrassing ones teach the most.

https://www.harvestkernel.com/core-grammar-drill/

The Four Reads is live in the Resources space.

Four phases, about 95 minutes, built to run on one syllabus you already have open:

Seed, 15 min. The Cold Read. Open your syllabus on a phone and read it as a nervous student. Collect evidence, fix nothing.

Sprout, 20 min. The Tone Swap. Rewrite the three coldest passages so the policy is identical and the reader is a person.

Grow, 25 min. The Policy Honesty Audit. Make your AI policy something a serious student can actually follow.

Harvest, 35 min. The Interview Paragraph. Write the sentence your student quotes in a job interview two years from now.

Interactive and printable, nothing stored on our servers, copy button at the bottom so you can take your answers with you.

https://www.harvestkernel.com/four-reads-syllabus-kit/

Post your Harvest paragraph when you get there. That one is worth reading out loud.

New kit: The Crossover Ladder.

Four rungs for taking on work outside your job title without pretending you have the training.

Seed, 10 min. Name the border. Which task, to which person, waiting on what.
Sprout, 20 min. Borrow the vocabulary. Ask AI what a senior person in that role asks before they start, then find the question you cannot answer.
Grow, 30 min. Build a throwaway. Make it badly on purpose, then have AI critique it against the standard in that field.
Harvest, 35 min. Show it to the owner. Ask what you got wrong, not whether it looks ok.

Rung four is the one people skip. Skipping it is what turns crossover into damage.

It is a working page, fill it in as you go and copy your answers out at the end. Nothing gets stored.

https://www.harvestkernel.com/crossover-ladder/

New in the vault: The Trust Shift Kit

A four phase SeedStacking kit to replace AI detection with design, for any faculty tired of accusing honest students on the word of a black box.

Seed, retire the detector and write a one sentence AI policy for one assignment. Sprout, require the Trust Receipt: the prompts, the output, and what the student changed. Grow, build a rubric that grades the judgment they added on top. Harvest, let them defend one choice out loud for two minutes.

Interactive and free: https://www.harvestkernel.com/trust-shift-kit/

When you reach the Harvest phase, post your two minute defense question back here so other faculty can borrow it.

New resource just went live: The Faculty AI Deploy Kit.

It is a four phase SeedStacking activity set that moves you from watching AI happen to grading it well. Seed, do your own next assignment with an AI tool once. Sprout, rewrite one assignment so AI use is required and visible. Grow, build a short rubric that grades the thinking, not the typing. Harvest, have students keep one AI assisted artifact they could show an employer.

No prior AI experience needed. The first rep takes about fifteen minutes.

Open it here: https://www.harvestkernel.com/faculty-ai-deploy-kit/

When you reach the Harvest phase, post your one paragraph employer framing back here so other faculty can borrow it.

New resource in the vault: The Classroom AI Boundary Card

The research says students and teachers already agree on where AI belongs. This one-pager turns that agreement into something you can post for a single assignment, plus a 15-minute first rep to close your own Readiness Gap this week.

Two parts: the boundary card (AI welcome here / AI stays out here) and a six-step first rep you run before your students do.

Grab it here: https://www.harvestkernel.com/classroom-ai-boundary-card/

New faculty resource: The AI Ready Course Redesign Kit

A free, interactive four phase kit to redesign a course for the AI era without burning it down. Audit, Reframe, Rebuild, Share.

Inside: the ten minute Offload Audit grid to find your risky assignments, an outcome reframe table to lift outcomes above what AI fakes, rebuild templates for AI resilient and AI integrated assignments, and a pilot and share one pager for your department.

Work it in one sitting or one move per week. Open the kit: https://www.harvestkernel.com/the-ai-ready-course-redesign-kit/

Run the audit on one assignment today and drop your reddest one in the Daily Discussion thread.