user-in-search-of-a-name/mike-willchen96
A 'Matter OS' workspace that turns each legal matter into a curated knowledge base with agent tooling and an activity feed.
This fork, from user-in-search-of-a-name, builds a "Matter OS" layer on top of Mike's existing project-and-matter model. The idea is to treat every matter as a structured workspace rather than a folder of documents: a curated knowledge base of typed entries - facts, parties, dates, clauses, positions, playbooks, and sources - paired with agent tools that can work against that knowledge, and an activity feed that tracks what's happening on the matter.
The handle itself tells you where this is: user-in-search-of-a-name hasn't settled on branding yet, and the fork is exactly one commit ahead of upstream. So this is a single, opinionated bet about how legal AI workspaces should be organized - curate the matter's knowledge once, then let agents and people work from a shared, typed source of truth.
If you're curious how the knowledge model is shaped, or where the agent tooling is headed, the GitHub repo is the place to look. It's early, but the direction is legible.
What's in it
- Matter OS workspace Reframes each matter as a structured workspace rather than a pile of documents - knowledge, tooling, and activity in one place.
- Curated knowledge base Typed entries - facts, parties, dates, clauses, positions, playbooks, and sources - give a matter a deliberate, structured memory.
- Agent tooling over the matter Agents are wired to work against the curated knowledge, not just raw documents.
- Activity feed A running view of what's happening on a matter, so the workspace stays legible as work accumulates.
Direction
knowledge-managementworkflow
Activity
Threads of work (detailed view)
user-in-search-of-a-name gives Mike a matter-long memory
A fork that turns the assistant from a per-chat helper into a system that remembers a matter across every conversation.