Rough observations, questions, sketches, clips, and provisional ideas captured close to the moment they are noticed — before interpretation hardens into explanation.
A development Fieldnote tracing Ana from its GraphRAG-oriented beginnings through governed knowledge formation, memory, Knowledge Spaces, Domain Packs, Workbench, canonical reference integrity, and cross-canon contracts.
An engineering Fieldnote on what Codex operating practices reveal about context entropy, working memory, tool exposure, handoffs, skills, and the emerging discipline of engineering the runtime around an intelligent agent.
A Fieldnote prompted by Bob Morse's experiment with Claude Code, examining what happens when deep domain judgment can increasingly become executable and what that may mean for the RPM Machine.
Enterprise Intelligence
Governed Autonomy
Still Resonating
Observations that continue to connect to current inquiry.
Something happened while putting The Book of Musewoods together that I did not expect. I thought we were documenting a system that had already become reasonably clear to me, gat...