Observed arrival · 2026-10-12
Howware: From know-how to done
An open proposal for putting personal rules to work through AI agents, using plain-language instructions, small exact tools, and data the user owns.
- For
- People who want to encode personal workflows for AI agents
- Worth noticing
- Its nine principles reserve judgment for plain-language rules, exactness for tested tools, and outward-facing actions for explicit approval.
Field notes
The architecture separates a readable Protocol from small programs for exact operations, shared Adapters, owned Memory with history, and a Conversation interface; the page explicitly says the agent itself is not one of the five parts. Linked artifacts include a Manifesto, Whitepaper, Specification, Template, and GitHub, alongside example labels such as finanzas-personales, todo, and project-memory. Its “Honest limits” section acknowledges that code remains necessary where exactness matters.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue