Projects
Mickey Graf publishes some of his tools as open source, each with a clear status and a record of what has been tested.
Runmill picks up a Linear issue that meets its rules and runs Claude Code or Codex on it in an isolated workspace. The exact candidate commit must pass the required checks and a separate review in a fresh context. Only then does Runmill's own code open a GitHub pull request and start CI. The agent cannot push, open a pull request or merge: deterministic code decides each of those steps.
The automatic merge step is experimental.
Runmill on GitHubctxlane keeps profiles apart for Claude Code and Codex, each with its own state and directory context. Its wrapper stores static secrets in the operating system's credential store, while sign-in and model requests go through the vendors' official CLIs. It works on its own, without Runmill.
Tested against simulated vendor CLIs. Real accounts, keyrings and billing are untested.
ctxlane on GitHubBefore you hire an engineering firm, it helps to see how it builds. Neither tool was built for a client, but both follow the rule we apply in client work: the agent proposes a change, and deterministic checks and people decide what happens to it.
Review the documentation, code, tests and controls yourself first. Runmill is a developer preview with experimental automatic merge, and ctxlane has been tested against simulated CLIs only.
Yes. Once the candidate commit has passed the required checks and the separate review, Runmill's code opens the pull request. The agent itself never pushes or opens one. If your rules require a person to submit every change, check this against them first.
Harness engineering
The setup that tells an agent how to build, test and change your code, kept in version control.
Coding agent adoption
Agents taken from a few pilot teams to the rest of engineering, with training and one shared setup.
Evaluations on your repositories
Tools, models or settings tested side by side on changes you have already merged, before you commit to one.
AI pilot to production
What has to be in place before a successful AI pilot goes live: regression tests, versioning and rollback, monitoring, an owner, a handover and a staged rollout.
Outcome versus trace evaluation
Why an agent run can finish without errors and still handle the case wrongly, and how to check business outcomes separately from traces.
Bring one piece of work you would like to improve. The call takes 30 minutes and needs no access to your systems or your data.
Booking runs on Google Calendar.
Prefer to write first? Leave your email and a few lines, and we reply with times.
Look for the draft in your email program
Send it from there, and an engineer replies in person with times for a 30-minute call. If no draft has opened, write to us at miki@cloudsail.com.