What's written down, and where.
Start with the question you have. Short answers live here; exact commands and flags stay next to the code so they cannot drift.
FAQ
Do I need tests, does my code leave the machine, why is it free, what BSL means, how long a run takes.
Read the answersThe method
The engineering argument underneath the tool — inventory, call-graph mapping, dependency strategy, behavioral verification, and how the rewrite engine works.
Read the methodCLI reference
Every command, every flag, and the JSON shape each one emits. Lives with the code so it can't drift from it.
Open the READMEFor agents
A machine-readable summary of what PyMolt is and which MCP tool to reach for. Point your agent at it directly.
View llms.txtBenchmark
Measured A/B against a control arm — what the tool costs an agent, what it saves, how long a run takes, and the five results that argue against us.
See the numbersOpen numbers
Published metrics — stars, catalog size, test suite — plus what isn't measurable yet and what is never collected.
See the numbers