--- title: Why I Use My Own Coding Agent | daily.dev description: A developer built his own coding agent, OneLoop, to control the workflow layer around LLMs rather than the models themselves. He uses it with a Codex... canonical: https://daily.dev/posts/why-i-use-my-own-coding-agent-kuxvdynas twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Why I Use My Own Coding Agent | daily.dev og:description: A developer built his own coding agent, OneLoop, to control the workflow layer around LLMs rather than the models themselves. He uses it with a Codex... og:url: https://daily.dev/posts/why-i-use-my-own-coding-agent-kuxvdynas og:image: https://api.daily.dev/og/posts/KUXVDynas.png og:image:alt: Why I Use My Own Coding Agent og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Why I Use My Own Coding Agent **[BirkeyCo](https://daily.dev/sources/birkey)** · 5 min read · 1 upvotes · 0 comments ## Summary A developer built his own coding agent, OneLoop, to control the workflow layer around LLMs rather than the models themselves. He uses it with a Codex subscription via OAuth, plus OpenRouter and OpenCode subscription support, and runs it inside an Emacs comint buffer with custom OSC 7501 status reporting. Session logs from OneLoop led him to build OneSearch, a lexical and semantic search tool over code and agent sessions, now available as a binary. He warns building a demo agent is easy, but he's spent close to a year hardening OneLoop with tests, and says the leverage is real but not free. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### Why would someone build their own coding agent instead of using Claude Code or Cursor? Building your own agent gives control over the workflow layer around the LLM, not the model itself, since LLMs remain stochastic regardless of the agent used. A self-built agent lets you choose and combine models and providers (Codex via OAuth, OpenRouter, OpenCode) without being locked into one vendor's interface, and integrate it directly into your existing editor, such as an Emacs comint buffer. _Weighing a custom coding agent against off-the-shelf tools is easier with real builder experiences on daily.dev._ ### How much effort does it take to maintain a custom AI coding agent for daily professional use? It takes close to a year of ongoing development to make a self-built coding agent reliable enough for daily professional work, according to one builder of a tool called OneLoop. Supporting multiple models and providers means handling different protocols, authentication, and failure modes, requiring continuous bug fixes and test coverage so changes in one path don't break another. _Developers scoping a custom agent project can gauge the real time cost through builder write-ups on daily.dev._ ## Similar posts on daily.dev - [OneLoop: Local First Coding agent written in Rust with nix tooling](https://daily.dev/posts/oneloop-local-first-coding-agent-written-in-rust-with-nix-tooling-xaun3gjss) · BirkeyCo · 0 upvotes · 0 comments - [My $0 local coding agent does more than $20 Claude Code because it never says no](https://daily.dev/posts/my-0-local-coding-agent-does-more-than-20-claude-code-because-it-never-says-no-6c0y9ns7k) · XDA Developers · 1 upvotes · 0 comments - [Open-source coding agents like OpenCode, Cline, and Aider are solving a huge headache for developers](https://daily.dev/posts/open-source-coding-agents-like-opencode-cline-and-aider-are-solving-a-huge-headache-for-developers-77xyw8rmy) · The New Stack · 4 upvotes · 0 comments - [How I write software with LLMs](https://daily.dev/posts/how-i-write-software-with-llms-jekylkhup) · Hacker News · 2 upvotes · 0 comments - [How OpenAI Codex Works](https://daily.dev/posts/how-openai-codex-works-ianegzsz8) · ByteByteGo · 5 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#llm](https://daily.dev/tags/llm), [#devtools](https://daily.dev/tags/devtools), [#claude-code](https://daily.dev/tags/claude-code) [View this post on daily.dev](https://daily.dev/posts/why-i-use-my-own-coding-agent-kuxvdynas) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Why I Use My Own Coding Agent","url":"https://daily.dev/posts/why-i-use-my-own-coding-agent-kuxvdynas","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/why-i-use-my-own-coding-agent-kuxvdynas"},"datePublished":"2026-10-26T16:53:14.926Z","dateModified":"2026-10-26T16:53:40.651Z","description":"A developer built his own coding agent, OneLoop, to control the workflow layer around LLMs rather than the models themselves. He uses it with a Codex...","image":"https://media.daily.dev/image/upload/s--HRgLpUt6--/f_auto/v1722860399/public/Placeholder%2003","thumbnailUrl":"https://media.daily.dev/image/upload/s--HRgLpUt6--/f_auto/v1722860399/public/Placeholder%2003","isAccessibleForFree":true,"articleSection":"BirkeyCo","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"BirkeyCo","logo":"https://media.daily.dev/image/upload/s--VCFLlkwb--/f_auto,q_auto/v1774960036/logos/birkey?_a=BAMAMiWQ0","url":"https://daily.dev/sources/birkey"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/why-i-use-my-own-coding-agent-kuxvdynas","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,llm,devtools,claude-code","timeRequired":"PT5M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"BirkeyCo","item":"https://daily.dev/sources/birkey"},{"@type":"ListItem","position":3,"name":"Why I Use My Own Coding Agent"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/why-i-use-my-own-coding-agent-kuxvdynas#faq","mainEntity":[{"@type":"Question","name":"Why would someone build their own coding agent instead of using Claude Code or Cursor?","acceptedAnswer":{"@type":"Answer","text":"Building your own agent gives control over the workflow layer around the LLM, not the model itself, since LLMs remain stochastic regardless of the agent used. A self-built agent lets you choose and combine models and providers (Codex via OAuth, OpenRouter, OpenCode) without being locked into one vendor's interface, and integrate it directly into your existing editor, such as an Emacs comint buffer. Weighing a custom coding agent against off-the-shelf tools is easier with real builder experiences on daily.dev."}},{"@type":"Question","name":"How much effort does it take to maintain a custom AI coding agent for daily professional use?","acceptedAnswer":{"@type":"Answer","text":"It takes close to a year of ongoing development to make a self-built coding agent reliable enough for daily professional work, according to one builder of a tool called OneLoop. Supporting multiple models and providers means handling different protocols, authentication, and failure modes, requiring continuous bug fixes and test coverage so changes in one path don't break another. Developers scoping a custom agent project can gauge the real time cost through builder write-ups on daily.dev."}}]} ```