{"slug": "we-built-mdcms-a-markdown-first-cms-for-teams-using-ai-agents", "title": "We built MDCMS, a Markdown-first CMS for teams using AI agents", "summary": "A developer has built MDCMS, an open-source, Markdown-first CMS that allows both human editors and AI agents to manage content through raw files or a database-backed interface. The tool includes a CLI that scans existing repositories for Markdown and MDX files, automatically inferring content types and syncing them into the CMS as drafts. MDCMS aims to solve the adoption problem for teams with existing content in repositories, enabling AI agents to safely work with content at scale while maintaining validation, permissions, and rollback capabilities.", "body_md": "Working with Markdown files for content is super convenient.\n\nAs a developer, it is hard not to like it. Content is just files. You can search it,\n\ndiff it, review it in PRs, move it around, and now also let LLMs bulk-manage it.\n\nNeed to add metadata to 200 posts or migrate content from one structure to another?\n\nFiles are very easy to reason about.\n\nBut this does not scale that well for editors. Most marketing and content teams do not\n\nwant to open a repo, understand folder structure, run commands, and wait for deployments\n\njust to change a page. They need drafts, preview, permissions, version history,\n\npublishing flows, rollback, localization, and all the other CMS things that are still\n\nvery much needed.\n\nSo the idea behind MDCMS was pretty simple: why not both?\n\nMDCMS is an open-source CMS that is database-backed and Markdown-first. The database\n\nstays the source of truth, so you still get the workflow and safety of a real CMS.\n\nBut Markdown/MDX stays the working format, so developers and AI agents can still work\n\nwith raw files when that is the better tool for the job.\n\nThe part that matters a lot for brownfield projects is adoption. A lot of teams already\n\nhave content sitting in their repository: blog posts, docs, landing pages, changelogs,\n\nMDX pages, marketing content. The problem is not starting from zero. The problem is\n\nturning what already exists into something editors can safely manage.\n\nWith MDCMS, the goal is that you can run:\n\n```\nmdcms init\n```\n\nAnd the CLI walks through the existing repository. It scans for `.md`\n\nand `.mdx`\n\nfiles,\n\ngroups content directories, detects locale patterns, parses frontmatter, infers content\n\ntypes, generates `mdcms.config.ts`\n\n, syncs the schema, and pushes the discovered content\n\ninto the CMS as drafts.\n\nThat is the adoption path I care about most:\n\nAfter that, the workflow looks roughly like this:\n\nFor example:\n\n```\nmdcms pull\n# update content, metadata, links, translations, SEO fields\nmdcms push --validate\n```\n\nThe interesting part is not \"AI writes blog posts.\" The interesting part is giving AI\n\nagents a safe way to work with content at scale, while still keeping validation,\n\npermissions, history, and rollback.\n\nFor us, AI-native CMS means AI can work across multiple layers:\n\n| Layer | What AI can help with |\n|---|---|\n| Content | Markdown/MDX, metadata, links, translations, SEO fields |\n| Configuration | schemas, environments, locales, project settings |\n| Codebase | adapters, components, validations, missing workflows |\n| CMS actions | eventually controlled actions like users, roles, and permissions |\n\nThat is the part raw files alone do not solve. And it is also the part many CMS\n\ndashboards are not great at, because the content is often trapped behind UI flows and\n\nproprietary APIs.\n\nMDCMS is our attempt to keep the useful parts of both models.\n\nFiles when files are better.\n\nCMS when teams need a CMS.\n\nIt is open-source and still early, so contributions are very welcome. If this problem\n\nsounds familiar, check the repo, try it on a small Markdown/MDX project, open an issue,\n\nor pick something from the roadmap.\n\nWebsite: [https://www.mdcms.ai/](https://www.mdcms.ai/)", "url": "https://wpnews.pro/news/we-built-mdcms-a-markdown-first-cms-for-teams-using-ai-agents", "canonical_source": "https://dev.to/jjablonskiit/we-built-mdcms-a-markdown-first-cms-for-teams-using-ai-agents-2715", "published_at": "2026-05-25 15:15:24+00:00", "updated_at": "2026-05-25 15:33:45.820094+00:00", "lang": "en", "topics": ["ai-tools", "ai-agents", "ai-products", "ai-startups", "ai-infrastructure"], "entities": ["MDCMS", "Markdown", "MDX", "CMS", "LLMs", "AI agents"], "alternates": {"html": "https://wpnews.pro/news/we-built-mdcms-a-markdown-first-cms-for-teams-using-ai-agents", "markdown": "https://wpnews.pro/news/we-built-mdcms-a-markdown-first-cms-for-teams-using-ai-agents.md", "text": "https://wpnews.pro/news/we-built-mdcms-a-markdown-first-cms-for-teams-using-ai-agents.txt", "jsonld": "https://wpnews.pro/news/we-built-mdcms-a-markdown-first-cms-for-teams-using-ai-agents.jsonld"}}