AI with MattLearn as you work.Member area

A useful content week.
Within the time you have.

Plan a newsletter and a short post from three notes. Check that the work fits the time you have and leaves time for review.

The task

A made-up workshop wants a newsletter and a short post. One person has three hours to write and check them. Can the plan fit?

Ask a question you can check.

Use these notes to plan a newsletter and a short post. We have three hours, including time to check the work. Say what each item is for, who will do it and when it should be checked and shared. Show which notes support it. Tell me what is missing. Do not make up results.

The download contains sample inputs and an authored explanation. It does not contain the full member skill instructions. Supply your actual context when using the corresponding workflow.

Read the notes before using the plan.

The authored brief

Optional: read the sample notes as a data file
{
 "label":"Authored teaching brief, fictional workshop",
 "timezone":"America/Chicago",
 "start":"2026-11-09",
 "end":"2026-11-13",
 "capacityMinutes":180,
 "owners":["Sample educator"],
 "channels":["Newsletter","Short post"],
 "sources":{
  "S1":"A skill supplies reusable instructions. The assistant still needs the relevant task context.",
  "S2":"Check generated claims against the supplied source. Separate suggested checks from checks actually run.",
  "S3":"For a first task, choose a small source excerpt you are permitted to share."
 }
}

The proposed week

The sample has a newsletter on Tuesday and a short post on Thursday. Both leave time for the writer to check the work first. The plan uses made-up notes and dates.

135 minutes production + 30 minutes review = 165 minutes. Fifteen minutes remain. These are proposed effort estimates, not observed productivity.

Read the full sample plan (optional data file)
  1. Check what each source supports.

    Compare what the assistant wrote with the notes you gave it. If it adds a fact that is not in your notes, ask where it came from. Remove anything you cannot support.

  2. Count the work, including review.

    The newsletter allows 90 minutes to write and 20 to check. The post allows 45 minutes to write and 10 to check. Images, video and sharing need extra time. Add that time or make the plan smaller.

  3. Verify the calendar and the people.

    Make sure the people doing the work have time for it. Check who needs to approve each item before it is shared. This example has not checked anyone’s real calendar.

Optional: run the extra code checks

Run the structural checks.

From the extracted content-plan folder:

node check.js
node --test --test-reporter=tap
Checked 2 items: 165/180 minutes; 15 minutes unallocated.
Dates, listed owners/channels, approvals and source IDs checked. Timezone context: America/Chicago.
Human review still needed: factual meaning, effort estimates, permissions and actual calendar availability.

The included plan is accepted and five tests pass, zero fail, exit code 0. Negative cases reject overload, invented source IDs, late approval, dates outside the week, unknown owners, duplicate IDs and invalid estimates.

Read the actual planning test output
TAP version 13
# Subtest: the supplied two-item plan fits all declared structural checks
ok 1 - the supplied two-item plan fits all declared structural checks
  ---
  duration_ms: 1.055903
  type: 'test'
  ...
# Subtest: over-capacity proposal is rejected
ok 2 - over-capacity proposal is rejected
  ---
  duration_ms: 0.181753
  type: 'test'
  ...
# Subtest: an invented source reference is rejected
ok 3 - an invented source reference is rejected
  ---
  duration_ms: 0.692797
  type: 'test'
  ...
# Subtest: late approval and out-of-week publication are rejected
ok 4 - late approval and out-of-week publication are rejected
  ---
  duration_ms: 0.265451
  type: 'test'
  ...
# Subtest: unknown owner, duplicate ID and invalid estimate are rejected
ok 5 - unknown owner, duplicate ID and invalid estimate are rejected
  ---
  duration_ms: 0.142429
  type: 'test'
  ...
1..5
# tests 5
# suites 0
# pass 5
# fail 0
# cancelled 0
# skipped 0
# todo 0
# duration_ms 37.337236

You still need to read the work.

The extra code checks can catch missing details and a plan that runs over time. They cannot tell whether the writing is good or whether a task really takes as long as you guessed. Read the plan, compare it with your notes and check that people have time to do the work. This example does not publish anything.

Follow the member content path

Bring the checking habit to your own task.

Use your own permitted material, make the intended outcome explicit and inspect the result. Membership adds full available skills, setup guidance and starter paths; these samples remain free.