Session Architect
Plan the session before you open the chat. Fill in what you want, what the AI needs to know, and how you'll know it worked. Get a formatted opening brief you can paste into Claude, ChatGPT, Gemini — anywhere.
No AI involved here — this is a thinking tool. The value is in forcing yourself to be specific before you start prompting.
fill in the goal to generate your brief
formats into a structured message you can paste into any AI chat
Most bad AI output is a bad brief
When an AI gives you something useless, the instinct is to blame the model. But most of the time, the problem started before you hit send. You gave it a vague goal, no context, and no way to know when it's done.
A structured opening message doesn't guarantee good output. But it makes bad output diagnosable — you can see which part of the brief the AI missed, instead of trying to fix a conversation that was never set up to succeed.
Why these specific fields
- Goal — Forces you to state what "done" looks like in one sentence. If you can't, the task is probably too big for one session.
- Context — Everything the AI can't infer. Stack, constraints, what you've already tried. Skipping this is the #1 cause of irrelevant output.
- Files — Tells the AI what to expect. "You'll see three files" works better than pasting everything at once with no roadmap.
- Criteria — Moves the conversation from "is this good?" to "does this pass?" Concrete success criteria catch more bugs than vague satisfaction.
- Escape — Permission to stop. Without this, AI models tend to push through with a bad approach rather than flagging that the task needs rethinking.
About the templates
The templates aren't prompts — they're starting points. Each one pre-fills the fields with placeholders you're meant to replace with your specifics. The goal is to save you from staring at blank boxes, not to give you a prompt to paste blindly.
Everything happens in your browser. No data is sent anywhere — you copy the brief yourself and paste it wherever you want.