Writing Review - Prompting Is the Door, Not the Destination

Executive Summary

The packet has a strong strategic correction: prompt engineering belongs early, but it should not become the identity of AI literacy. The “door, not destination” framing is durable and helps avoid both prompt-marketplace hype and condescending beginner instruction. The main sequencing risk is now resolved in source: this module has its own slug and sits before the future Conversation module.

Must Address

• Keep the new slug stable: prompting-is-the-door-not-the-destination. Do not reuse the old conversation-oriented slug for this module.
• Define the planned Conversation and Context modules soon enough that this module’s bridge points somewhere real. The bridge is conceptually strong, but without follow-on source notes it may feel like a promise without a path.
• Keep the two safety gates visually separate in learner materials. The packet is right that input safety and output responsibility should not collapse into one checklist.

Should Address

• Add a beginner-friendly explanation of “context” before asking learners to provide it. The word is accurate but can become vague.
• Keep the lab tasks low stakes in every demo. Small-business communication can become consequential quickly if it involves legal, employment, medical, refund, or pricing claims.
• Test the lab with slow typists and observers. The accessibility guidance is good, but a 60-minute session can still become typing-heavy.
• Decide whether “Prompt Engineering” remains an internal category label while “Prompting Is the Door, Not the Destination” is learner-facing.

Nice to Have

• Create printed prompt cards for the three demo examples.
• Add one “I got a bad answer; now what?” troubleshooting block to the facilitator guide.
• Add a short “when to restart” teaser without teaching the whole Conversation module.

Structure and Flow

The sequence is strong: ordinary help-seeking, broad prompt, four handles, progressive demo, lab, safety gates, and bridge. I would avoid introducing magic-prompt culture too early in a live session; it works better after learners have experienced the repair loop.

Accuracy and Claims

No major accuracy issues. The durable-versus-changeable rule is important and should prevent stale vendor guidance. Avoid any claim that one prompt model works equally well across all models, accounts, and use cases.

Clarity and Concision

The packet is clear as source material. Learner-facing copy should be much shorter and should use examples before abstraction.

Suggested compression:

Start with what you want, add what the chatbot needs to know, say what kind of answer would help, and name anything it should avoid. Then inspect the answer and keep steering.

Section-by-Section Notes

Core Position

Strongest part of the packet. Preserve the distinction between prompting as a useful skill and judgment as the true north.

Foundational Prompt Model

Good lightweight model. Keep it flexible; do not turn it into a branded acronym unless a pilot shows learners need one.

Repair Loop

This should be the heart of the live experience. It normalizes imperfect starts and reduces beginner shame.

Two Safety Gates

Excellent distinction. This also creates a good bridge to privacy and verification modules.

Magic Prompt Culture

Useful and respectful. The tone should be “some of these tools are real, but later,” not “people on the internet are silly.”

Acceptance Criteria

Good pilot bar. The “two people outside the core design group” requirement is practical and should stay.

Strengths

• Clear non-hype framing.
• Strong beginner empathy.
• Good standalone value.
• Good bridge into Conversation and Context.
• Strong safety distinction between inputs and outputs.
• Good v0 boundary around lightweight facilitation instead of heavy slideware.

Edit Tracker

#	Priority	Item	Section / Line(s)	Status
1	High	Keep Door module separate from the future Conversation module	Module Lineage	[x]
2	High	Create placeholder source notes for Conversation and Context modules	Bridge	[ ]
3	High	Keep input and output safety gates visually separate	Learner Handout	[ ]
4	Medium	Define context in beginner language	Prompt Model	[ ]
5	Medium	Add troubleshooting block for poor first answers	Facilitator Guide	[ ]
6	Medium	Test with slow typists, observers, and casual users	Testing Tasks	[ ]

---

Source: Feedback/prompting-is-the-door-not-the-destination/2026-07-10 09-21 - codex - writing-review.md