Broadcast to agents
The ninth channel: agents.
Coding agents check what a dependency shipped before upgrading it. Yanib serves every story in machine readable formats, markdown for LLMs, JSON for tools, MCP for agents, so the answer is one request away, no scraping required.
Free forever. No credit card.
How it works
Three steps. Two minutes.
Connect
Connect a repo and publish. Machine readable serving is not a setting, every public story page gets it automatically.
Review
Agents get exactly what humans approved. The markdown at llms.txt is the reviewed story, not raw commit logs.
Broadcast
Agents read /stories/your-repo/llms.txt as clean markdown, hit the JSON API, or connect to the public MCP server, read only, no account.
Why the broadcasting layer
AI agents is one spoke of nine. Every channel gets the same reviewed update, from one publish.
- Per story page llms.txt: every story as clean, token efficient markdown
- Public MCP server, agents connect without an account or key
- Structured JSON via the v1 API for tool builders
- A site wide /llms.txt index so agents can discover all of it in one fetch
Frequently asked questions
Why would an agent read my release stories?
Dependency upgrades are the obvious case: before bumping your library, an agent reads what changed between versions. Reviewed release notes answer that far better than a git log.
What is MCP?
Model Context Protocol, an open standard for connecting AI assistants to tools and data. Yanib exposes a public, read only MCP server, plus an authenticated one for drafting and publishing your own releases from an agent.
Do I have to configure anything?
No. If your story page is public, its llms.txt, JSON API, and MCP access exist automatically. Machine readability is a property of the platform, not a plan feature.