Blog — Intrane.fr (Page 3)

Product · July 2026

The Ratchet — How My AI Fleet Merges Its Own PRs

An AI agent that opens PRs is easy. An AI agent that safely merges them is hard. I built a ratchet — four levels of escalating cleverness that let my merger self-unblock 80% of gated PRs before it ever asks me to look. Last night it merged 5 PRs in 7 minutes while I slept.

#automaintainer #ai-agents #devtools #automation #am-operator

Product · July 2026

A 100 KB Git-Like VCS That Beats Git at Its Own Game — Why I Built lume

lume is a 100 KB native version-control system written in machin/MFL. It is git-like, not git-based, and it now benchmarks faster than git for add+commit on 110 small files.

#vcs #git #machin #mfl #developer-tools

Product · July 2026

Your CLI Should Update Itself

I have 6 servers running remotecmd daemons. Last week I spent 20 minutes manually updating each one to the latest version. That's stupid. So I implemented the cli-update-spec: rcmd update downloads, verifies, smoke-tests, and atomically swaps the binary. No SSH, no installer, no downtime. Here's how it works and why every CLI should do this.

#cli #self-update #agent-first #remotecmd #open-source #devtools

Product · July 2026

The Agent That Signed Up and Paid for Its Own Analytics

A live, unedited transcript: an AI agent discovers vigie's API from llms.txt, creates an org, hits its free quota, reads the HTTP 402, links a prepaid péage wallet, and pays for more events — real cents on the live rail, zero humans on either side.

#vigie #peage #x402 #agent-first #agent-economy

Product · July 2026

GitOps for People with Three VPSes and No Kubernetes

My control plane deploys itself: push → build on the box → a human gate on email → atomic swap → health-or-rollback. One approval, one cent, zero LLM calls — and the pipeline lives in git, resolved fresh at execution. How it compares to ArgoCD and Jenkins, and the one twist neither can copy.

#gitops #deploy #roam #machin #argocd #jenkins

Product · July 2026

I ran a coding agent on my own CPU — LLM engine written from scratch in a language you've never heard of

A coding agent completed a task on a low-power CPU with no cloud and no PyTorch — the entire LLM engine written from scratch in machin (MFL). Including the moment the 1.5B model tried to cheat the test, and how I caught it.

#machin #LLM #self-hosted #AI #from-scratch #machin-anvil

Product · July 2026

Web Analytics With No Dashboard (But There Is a Globe)

vigie is cookieless web analytics built for agents: JSON verbs instead of dashboards, on-demand HTML reports, and a living 3D globe with city-level dots. Full metric parity with the dashboard-first tools — geo to city, journeys, vitals, errors, heatmaps — in one MIT binary, hosted tier metered per-use.

#vigie #analytics #agent-first #machin #peage

Product · July 2026

grange: A Database Your Agent Can Pay For

I built a document database in machin — my language for AI agents — and benchmarked it honestly against SQLite: faster on every indexed workload, some by 1000x. Then I gave it the missing piece: an agent signs up with a payment wallet, gets an isolated namespace, and pays €0.15/GB/month. No console, no card form, no human in the loop.

#machin #grange #database #agents #peage #dogfooding

Product · July 2026

POST a Spec, Walk Away, Get a Webhook: the roam Stack

Every agent dies with my terminal — so I built the stack that doesn't: roam (an agent you scp to a box), roam-panel (one-tap approvals on your phone), and roam-hub (a control plane that schedules, meters, and bills autonomous runs). All MIT, all single binaries, all machin.

#agents #roam #machin #autonomous #peage

Product · July 2026

Four Specs That Make Any CLI Agent-First

I ship a lot of CLIs — grepapi, mago, remotecmd, chatsnip, automaintainer. Each one needed the same four things: a way for agents to read its output, learn its model, report bugs, and update itself. So I wrote four tiny specs that any CLI can adopt in 30 minutes each.

#cli #agent-first #open-source #specs #devtools