I tested whether Claude Code follows decisions from earlier chats

Avinash Jetwani — Mon, 05 Oct 2026 14:02:38 +0000

I got tired of re-explaining my project to Claude Code every session.

We'd decide something in one chat. "We use Postgres." "Don't touch the public API." "That migration approach failed, and here's why." A few sessions later, I'd be typing it again.

CLAUDE.md fixes this if you keep it updated. I didn't.

So I built jevmem.

What it does

You decide something in a Claude Code chat.
After each message, Jev (TypeSafe AI's decision model) checks if it's worth keeping: a decision, a rule, a bug, an approach that failed.
If it is, jevmem writes one line to JEVMEM.md in your repo.
Next session, the lines that matter for your prompt go back to Claude.
Change your mind, and the old line is crossed out.

Your team gets the same file through git.

Does it work? I measured it.

24 tasks in 3 small projects, 3 runs each. Real Claude Code sessions with claude-sonnet-5. Each task has a right answer that depends on a project decision saved earlier.

Three setups: no project memory, jevmem, and the same lines written by hand into CLAUDE.md.

	No memory	jevmem	CLAUDE.md
Followed the project's decision	28/72	66/72	67/72
Tried a change the project forbids	10/18	0/18	not run
Repeated an approach that had already failed	3/15	0/15	0/15

So jevmem is about as good as a hand-written CLAUDE.md. The difference is that you don't write it.

Where CLAUDE.md won: a convention nothing in the prompt points to, a rule for every user-facing string. CLAUDE.md got it 3 of 3 times, jevmem 0 of 3. Rules every task must follow still belong in CLAUDE.md.

What to know

It needs a TypeSafe API key.
Message text goes to TypeSafe to be scored. Common secrets are scrubbed first. No telemetry.
Automatic in Claude Code. Codex and Cursor use the same file over MCP.
These are my tasks and my projects. The scripts and results are in the repo, so you can check them.

Try it

Claude app → Plugins → Discover → jevmem → Add
npm i -g jevmem
jevmem key, then jevmem enable in your project

Open source, MIT. Built on TypeSafe AI's Jev.

Repo: https://github.com/Avinash-jetwani/jevmem · Full results: https://avinash-jetwani.github.io/jevmem/results/

I'd love to hear where it saves the wrong thing.

Written with help from Claude. The numbers are from my own test runs, and the results files are in the repo.

I built a Polymarket idea tester that will kill your strategy (paper-only, no wallet)

Avinash Jetwani — Fri, 25 Sep 2026 13:51:41 +0000

I keep seeing the same ask: how do I test a Polymarket idea before I put money on it?

Most answers point at one of three things. A live simulator with fake cash on live odds. A Python backtest notebook. Or a dashboard that never quite says stop.

I wanted a fourth option: type the idea in English, run it on resolved history, and get a verdict I could read without translating a log. Keep. Kill. No edge. Inconclusive.

That product is OddSage — https://oddsage.com

What it is
OddSage is paper-only Polymarket research. No live trading. No wallet. No copy-trade execution.

Two ways in:

Paste a Polymarket URL. You get a Market Dossier — a research page for that market, not an order ticket.
Type the idea in English. The builder drafts a strategy. You read the draft. Save and run.
The loop is: English idea → draft → Save & run → verdict.

What the verdicts mean
Keep: the paper test left something worth holding. Still paper. Not a live order.

Kill: it did not survive.

No edge: it ran and still looked like noise.

Inconclusive: not enough evidence. That is a stop, not a soft keep.

A tool that only ever says “interesting” is a mood board. I wanted one that would kill an idea.

What it is not
It is not a live virtual wallet. If you want to practice clicking on live odds with play money, use a simulator. That is a different job.

It is not a Python SDK. If your edge already lives in a notebook with L2 replay, keep that workflow. OddSage is for people who have a sentence and want a keep/kill without opening a repo.

It will not copy another trader onto your book. It will not place the trade if the verdict is keep.

A public kill
I am not going to invent a win rate for a launch post. Here is a finished public paper-test that did not get a pass:

https://oddsage.com/strategy/259bc179-04dc-4e6d-816d-59d8f3cb52cc

If a tester never shows a kill, it is selling hope.

Why I am posting this on DEV.to
Because the SERP for “polymarket strategy tester” and “test polymarket strategy” is full of Python tutorials and sim waitlists, and the category still needs a clear paper-only option that admits failure.

Tagline I keep using: Test your Polymarket idea before you risk real money.

Product pages if you want the job by name:

https://oddsage.com/polymarket-strategy-tester
https://oddsage.com/test-polymarket-strategy
https://oddsage.com/polymarket-paper-trading

I’m Avinash Jetwani. On X: @jetwaniavinash

If it dies, it dies on paper.

DEV Community: Avinash Jetwani

I tested whether Claude Code follows decisions from earlier chats

I built a Polymarket idea tester that will kill your strategy (paper-only, no wallet)