Mendral

Blog

Product

Automating the Engineering Work That Isn't Your Product

The DevOps work every team has, Mendral already does. The work only your team has, the chores no vendor ships, you can now hand to an agent by describing it in Slack. How Automations work, and the seven that Blaxel runs in production without writing a line of code.

Sam Alba
Industry

What Flaky Tests Cost a 75-Engineer Team

A 75-engineer team loses about $375,000 a year to flaky tests in developer time alone. Here's the math, the CI compute bill, and the costs you can't put in a spreadsheet.

Sam Alba
Industry

You should not update your dependencies in 2026

A brief (irreverent) history of software supply chain security, and what to do about it in the age of AI.

Olivier Gambier
Product

Supply Chain Attacks Don't Wait for CVEs

By the time a CVE exists, the bad version is already in thousands of CI pipelines. Here's the last year of npm and Actions attacks, why scanners are reactive by design, and the supply chain agent we shipped to catch dependency changes at the PR.

Sam Alba
Engineering

How We Know If Our Agent Is Right

36,564 investigations, no ground truth, and the eval signals we actually trust.

Sam Alba
Engineering

The Agent Harness Belongs Outside the Sandbox

Two architectures for running agent harnesses, the tradeoffs between them, and how we make skills and memories work when the harness isn't local.

Andrea Luzzardi
Engineering

Same LLM, Different Agent: What Changes When You Specialize for CI

Same models, different tokens. What changes when you build an agent harness specialized for CI instead of general-purpose coding.

Sam Alba
Engineering

We Upgraded to a Frontier Model and Our Costs Went Down

We switched to a frontier model and our costs went down. Here's the architecture that made it possible.

Andrea Luzzardi
Engineering

We use Claude Code daily. We still built our own CI agent.

How Mendral closes 16,000 CI investigations a month: three Anthropic tiers, Firecracker microVMs, durable execution on Inngest, and a custom Go agent loop.

Sam Alba
Engineering

LLMs Are Good at SQL. We Gave Ours Terabytes of CI Logs.

We gave our AI agent a SQL interface to billions of CI log lines in ClickHouse. How we ingest, store, and query 1.5 billion log lines a week.

Andrea Luzzardi
Industry

What CI Actually Looks Like at a 100-Person Team

575K CI jobs, 1.18 billion log lines, 33 million test executions in one week. What we learned building an AI agent for PostHog's CI at scale.

Sam Alba