The latest articles on DEV Community by Maxime Houle (@maximehoule). https://dev.to/maximehoule https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4038387%2F9c714ae0-1eaf-4f55-a2ac-ccffa6fe0d79.png https://dev.to/maximehoule en Maxime Houle Mon, 20 Jul 2026 15:14:39 +0000 https://dev.to/maximehoule/i-built-an-open-source-ci-for-ai-agents-and-here-is-why-i-went-open-source-47nn https://dev.to/maximehoule/i-built-an-open-source-ci-for-ai-agents-and-here-is-why-i-went-open-source-47nn <p>Hi, my name is <a href="https://x.com/themaxthule" rel="noopener noreferrer">Max</a>. I built <a href="https://github.com/seldonframe/reelier" rel="noopener noreferrer">Reelier</a>, an open-source tool that gives AI agent workflows the thing normal code has had for decades: snapshot tests and CI.</p> <p>Here is the itch. </p> <ul> <li>My coding agent kept re-deriving the same tool-call workflow on every run. </li> <li>Fetch this, transform that, post the result. </li> <li>Same tokens, same seconds, for a task with exactly one correct answer. </li> </ul> <p>And every so often it did it a little differently, and nothing caught it. That second part is the scary one.</p> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fbkh2u8b4ja95rqnygg2n.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fbkh2u8b4ja95rqnygg2n.png" alt=" " width="800" height="453"></a></p> <h2> Why open-source it </h2> <p>Agent tooling is flooded right now. New closed tools ship every day, and most of them ask you to trust a black box with your workflows and your keys. </p> <p>Open-source is how a small tool earns trust in that noise: you can read exactly what it does, run it on your own machine, and keep your data as plain files.</p> <h2> What it actually does </h2> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fczp0dmo36dp2x0dnp0ip.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fczp0dmo36dp2x0dnp0ip.png" alt=" " width="800" height="286"></a></p> <ul> <li>You record one agent run that worked. </li> <li>Reelier compiles it into a readable SKILL.md file, a recipe with typed inputs, typed outputs, and an assertion on every step. </li> <li>Then it replays that file deterministically: no model call, 0 tokens, byte-identical, in milliseconds. </li> <li>Every replay is a receipt. </li> </ul> <p>And reelier diff compares two runs and exits 1 when a step drifts, so a silent regression fails your build instead of reaching production.</p> <h2> The honest part </h2> <p>It only replays deterministic tool calls, HTTP and MCP. </p> <p>It does not replay file edits or free-form reasoning, and it never fabricates a step it cannot reproduce. It is for the deterministic half of what agents do. </p> <p>I say that up front because the version that would go viral (record my whole agent) is the version I will not claim, since it would not be true.</p> <p>Numbers, from the benchmark in the repo: 1000 of 1000 replays byte-identical, 0 tokens per replay, about 50x cheaper and 59x faster than the agent re-running the same task.</p> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fbcg47avefcsxj3vc2l0p.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fbcg47avefcsxj3vc2l0p.png" alt=" " width="800" height="375"></a></p> <p>It is AGPL-3.0, BYOK, and runs entirely on your machine.</p> <p>Any help you can give me would be super tremendous! 🥹<br> I would be happy to get a star if possible ❤️<a href="https://github.com/seldonframe/reelier" rel="noopener noreferrer">https://github.com/seldonframe/reelier</a>. </p> ai opensource showdev webdev