Skip to content
FlintFlintby Torch
← Level 6Builder · L6

Demo Day: one full cycle

Build-along · ~20 min at your own keyboard

This is the gate. Everything you built now runs as one cycle: a trigger starts it, the roles hand off, the result lands where you said, you review it, and the log records it. Be honest about how the hand-offs happen. Most no-code tools today can't make one Project or GPT start another, so in the no-code route you are usually the courier: you paste each hand-off into the next role, unchanged. That's a real, working system — call it manually operated, and record it that way. Only call it automatic if a tool actually starts each role and passes the hand-off without you. Pick a need that really recurs in your work. You show evidence, not a description: timestamps, the output, the log lines, and one thing that went wrong and how you handled it. This lesson is reviewed by a person, at the Tuesday capstone or by Steve. It is a step past the single-system capstone in the Builder track: that one proved one system runs. This one proves the whole set runs together.

Ladder · Level 6 gateThe long version: Demo day: a full cycle ↗

Human review · Class week 12 (Demo Day)

Steve Gustafson or the class lead: A full cycle end to end, with timestamps. Reviewed against the rubric above and the Level 6 done check.

The rubric they use →

The coach's check below is not this review. This lesson stays provisional until a person has looked at it with you.

Do this

0/7

Checkpoints

  • ▸Every role ran in order and each left a Log line.
  • ▸You've written down whether the hand-offs were automatic or passed by you.
  • ▸A result sits where the operating page says it should.
  • ▸You made a recorded review decision on it.
The three ways this goes wrong
  • A role ran but its hand-off was empty or malformed. — Open the hand-off, compare it to the format, and tighten the format example in that role's instructions. Run again.
  • The scheduled run did not fire. — Check the app is open or the machine is awake, the time zone, and that the task is enabled. Run it once by hand and say so in the evidence.
  • The output is fine but you cannot show when it ran. — Pull the timestamps from the run history or log files now. If none exist, add a timestamp line to the Log format and run again.

Dated facts in this lesson: Claude Cowork scheduled tasks ("Scheduled" in the sidebar) run a saved prompt in the cloud with the laptop off. They need Pro or above (researched Oct 4, 2026, not yet confirmed by a person: check the source before you rely on it); ChatGPT has agent, Projects (project-only memory, files, instructions), memory toggles in Settings > Personalization, and Scheduled Tasks at chatgpt.com/schedules. Workspace agents ("ChatGPT Work", July 2026) are on Business and Enterprise. Agent quotas: Plus 40 messages a month, Pro 400. Secondary source (researched Oct 4, 2026, not yet confirmed by a person: check the source before you rely on it); Gemini Spark is a 24/7 agent with Skills and Schedules. Schedules need Pro or Ultra, personal accounts only, up to 50 schedules. A Mac version exists. Spark is a beta (researched Oct 4, 2026, not yet confirmed by a person: check the source before you rely on it); Microsoft 365 Copilot has Notebooks, Agent Builder agents, Pages, Copilot Cowork and Researcher and Analyst agents. Personal Copilot has Memory and Personalization settings, and Notebooks with Microsoft 365 Personal, Family or Premium. Secondary source (researched Oct 4, 2026, not yet confirmed by a person: check the source before you rely on it); Grok has Projects (files and instructions), "Personalize with Memories" under Data Controls, and Automations on a paid tier. SuperGrok Heavy ($300) runs parallel agents. Secondary source (researched Oct 4, 2026, not yet confirmed by a person: check the source before you rely on it); Perplexity has Memory (also on Free), Spaces (files and instructions; Pro allows 50 files per Space; not on Free), Tasks (at most hourly) and Computer on Max. Secondary source (researched Oct 4, 2026, not yet confirmed by a person: check the source before you rely on it).

Share what happened

Paste the evidence set: a timestamp per role, an output sample, the Log lines, how the hand-offs moved (automatic or by you), your review decision, and the one thing that went wrong.

Strip API keys, passwords, and tokens before you paste. Swap them for [KEY].

🔒 Don't paste confidential, personal-health, or payment info.