Lemma calculates real Axiom fees offline: open-source Rust app for Windows

⚡ The gist in 5 seconds

• The raygullcrease/axiom-launch-replay repository now hosts Lemma — a desktop tool for Axiom traders: offline replay of recorded launches and a full breakdown of per-trade costs.
• Installs as a single Lemma-setup.exe build on Windows, no login, no extensions, nothing to connect.
• The code is unsigned — SmartScreen will complain on first launch, so you'll have to click "More info" → "Run anyway".

🔍 What happened

A project called Lemma has appeared on GitHub — a Rust utility under the MIT license, 44 commits, around 200 stars and 62 forks. The author describes it as an independent companion for those trading on Axiom. It runs locally, installs on Windows, and the link to the releases is in the repository.

How it works. Lemma queries the chain through your own RPC and free public APIs, and every new launch gets saved to a folder on your disk. You then write rules — when to buy, when to sell — and replay weeks of launches in seconds, comparing the curve before and after costs. A separate tab checks a token by a single pasted address: mint and freeze authority, the real pool, same-slot bundles, holder concentration, the dev's past launches, copied names and images, and X accounts renamed within a week. You can also paste a public address and get a breakdown of every trade with its actual price, plus CSV export for Koinly and Awaken. There's also a local list of browser extensions that read trading pages.

Lemma doesn't execute trades. Everything it currently offers is backtesting, paper trading, and checks. An optional exit manager with spending limits and a transaction allowlist is on the roadmap.

🧠 Our take

The first thing that stands out: the author hits a very specific pain point. Terminals usually tell you what to buy. How much you're paying in fee tiers, priority fees, tips, and slippage — almost nobody shows that. It looks like someone simply got tired of calculating it by hand and built a tool for themselves: MIT license, zero logins, zero sending your address to someone else's server.

What this changes in practice. For a wallet holder, the difference between "the strategy works" and "the strategy only works on paper" is precisely the cost line. Now it can be calculated locally, for free, and without a subscription. Our assessment: the main value here isn't the backtesting but the tax CSV and the token check — those are exactly what people usually charge money for.

Who wins: those who trade themselves and don't want to hand their trade history over to a SaaS. Who loses — in our experience, paid terminals with analytics subscriptions. But we won't loudly bury them: they have support, SLAs, and signed builds, while a repository with 44 commits has none of the three so far.

⚠️ What we don't know

There isn't a single independent measurement. Exactly how much Lemma counts for priority tips and slippage, how many launches the local folder can hold, how much memory the "weeks in seconds" replay will eat — the README says nothing about any of this. The thresholds for the token checks aren't disclosed either: "same-slot bundles" sounds nice, but under what condition the flag is raised, the reader isn't shown. The unsigned exe is separately concerning: you can't verify that the binary was built exactly from this code. And 200 stars isn't a reputation yet — it's just 200 stars.

Prediction: within a quarter, either a fork with a signed build and CI will appear, or the project will quietly stay at its current 200 stars. I'm betting on the latter — code signing takes money and hassle, and there's no maintainer with a team visible here.

Disclaimer: not officially confirmed — the accuracy of the calculations and the safety of the build have not been verified by anyone outside; all details come from the README of a single repository.

Source: GitHub