The Threads Memory System

Never Re-Explain Yourself to Your AI Agent Again

Give your AI agent persistent memory across conversations. Not the kind that saves little tidbits of information about you. Not a complex external memory system. Just the ability to ask "where are we at with that project?" two weeks later — and your AI agent not skipping a beat.

threads/q3-launch.md
1# Q3 Product Launch
2
3## Status
4Done: Sales page copy finalised. Checkout flow tested.
5 Email sequence drafted (5 emails). Design approved.
6Next: Schedule launch emails. Set up affiliate links.
7
8## Decisions Made
9- Price point: $197. One-time, no upsell on checkout.
10- Launch window: 5 days. Cart closes Sunday midnight.
11- Affiliates get 40% — invite list confirmed (8 people).
12
13## Open Questions
14- Bonus: include live call or keep async only?
15- Waitlist email: send Monday or Tuesday morning?
16- Checkout page headline — still testing two versions.

Two folders. That's the whole system.

Most people assume AI memory has to be complicated — vector databases, retrieval plugins, elaborate setups that need constant maintenance. It doesn't.

The Threads Memory System is embarrassingly simple: a threads folder where each ongoing project gets its own running context file, and a daily-logs folder that tracks what happened in recent sessions. Before each conversation, your AI agent reads what it needs to know. After each session, it writes what happened.

No engineering. No maintenance. Just the ability to ask "where are we at with that project?" two weeks later — and get a real answer.

What's in the Toolkit

A setup guide your AI agent installs for you — hand it to Claude, say "set this up," and it handles everything: the folders, the skills, the memory instructions. You tick off each step as your AI agent works through them.
Project memory that survives every conversation — the threads folder gives your AI agent context on every ongoing project. Open a new conversation and it already knows where things stand, what decisions were made, and what comes next.
Short-term memory for recent sessions — the daily-logs folder tracks what happened in recent sessions. Your AI agent reads the last few days' logs at the start of each conversation, so "the other day we were working on..." actually works.
Two pre-built memory skills — save-thread and save-log are already written, tested, and ready to install. Just say "save the thread" at the end of a project session and your AI agent handles the rest.
Optional nightly automation (Claude Code) — if you want the daily log to build itself without thinking about it, the toolkit includes a script that runs at 9pm every night, reads your AI sessions from the day, and writes the summary automatically.

Give Your AI Agent a Memory

Install it once and your agent will know your projects, remember your decisions, and pick up where you left off — every time.

No spam. Unsubscribe any time.

You're in!

Check your inbox — your toolkit is on its way.

Ev Chapman
Ev Chapman helps knowledge professionals get things done in partnership with AI. She teaches through her newsletter, YouTube channel, and the College of Knowledge membership.