Post

Interesting looking open, model-agnostic platform for collaborating with agents as a team and accumulating multi-modal artifacts and logbooks in a single place. @cahut.bsky.social cahut.cloud

cahut.cloudCahut: the home for everything your AIs makeDocuments, slide decks, dashboards and small apps live online as living pages, in your house style, on your own server, whichever AI made them. Cahut is in early development.

26/10/26, 3:50 pm0 reposts0 likes

Catalyst Cooperative

@catalyst.coop

A slideshow style presentation trying to imagine better processes for teams to plan and collaborate with agents up front, rather than only coming together in review at the end of the workflow: maggieappleton.com/planning-age...

maggieappleton.comPlanning with Agents: Divided Worlds, Boundary Objects, and Thicker InterfacesWhy we need richer, thicker interfaces and better boundary objects for collaborative planning with agents

Catalyst Cooperative

@catalyst.coop

A good convo with Shreya Shankar, a new professor (starting in 2027) at CMU on the dbt podcast about whether and how and under what circumstances data agents can really do analysis. Spoiler-alert: having clean, well-normalized tables & rich metadata helps a lot! roundup.getdbt.com/p/can-data-a...

roundup.getdbt.comCan data agents do real analysis?Carnegie Mellon Professor Shreya Shankar on why data agents fall short at real analytics work, what a semantic layer changes, and how to evaluate LLM outputs you can trust.

Catalyst Cooperative

@catalyst.coop

Interesting manifesto / set of computing principles 1. Computing is inherently social. 2. Therefore, our use of computers should always account for community. 3. Accounting for community entails concern for other people. 4. Concern puts a person’s agency before their utility. destructured.net/casa

destructured.netComputing As Social Activity