for agents, and for anyone who prefers text

the site as markdown.

every page of openthings.live as a plain .md file, generated from the pages themselves. open, readable, nothing hidden.

/markdown.md, the whole site /llms.txt, the index

/markdown/teaser.mdthe teaserwhat openthings is: open standards for human things, agents as the first interface, where our agents and context live, the two eras (two ais and a person carrying the message, then agents on both sides), what dies next (the ui, software and communication as we know them), the OHAN network, the six problems, the fixes on offer, why now (the launches that make it possible), what is coming, and the founder notethe page →
/markdown/why.mdthe long takewhy openthings exists, with data: the first and second order of ai, where the money goes, the five things that will make ai elevate, why ai becoming the base layer changes what gets built, the hard problems everyone is working on and where openthings stands, and the history of the connecting layersthe page →
/markdown/collapses.mdwhat collapsestake 03: what falls away once people adopt openthings, an open network protocol and agents on both sides, and the far better things that replace it, one by onethe page →
/markdown/agent-web.mdthe agent webtake 04: the web was built for eyes and the next one is built for agents. the visit fades but the question does not, search splits into finding, answering and ordering, every agent has a person or an organisation behind it, formats for the eye fall and shapes for agents rise, much of software becomes templates, the pieces being built today (mcp, a2a, NANDA, aweb, x401, agentid), and the open questionsthe page →
/markdown/takes.mdthe takesthe founder's writing in one place: the long take, the things, what collapses, and the agent webthe page →
/markdown/things.mdthe thingswhat openthings builds, in three directions: open software your model runs (opensoftware, small agents, solo ops), agents with a human behind them on the network (openproof, openDM, openhiring, any agent that joins), and the protocols (OHAN: key, name, card, wire, trust; compared with mcp and a2a). how they fit from the model up, why agents on both sides make the app in the middle redundant, what collapses when the agent comes before the system, and eight human problems mapped against itthe page →
/markdown/next.mdwhat is nextthe protocol and the work ahead: OHAN, intent sharing with agents, open source found for you, trust as the floor, and what gets openedthe page →
/markdown/connect.mdconnectwho should reach out: builders of local and voice models, agent to agent systems, data sharing between agents, personal agents and desktop hacksthe page →