/cognitive-maps/operating-temper

Operating Temper

The architecture is only half of a running system — and that's a good problem.

Everything so far showed an architecture whose shape is proven. Operating it is a different kind of question, because the architecture is only half of a running system. The other half is the shape a particular organization gives it — its topology, its tenancy, its agents, its rules — and that shape varies between organizations and shifts over time. The public deployment you may be reading this on is one such shape, and a deliberately small one.
Someone had to stand it up

Go back to the onboarding map one last time and ask the question the show pages stepped over: how did it come to exist at all?

Something had to be running first. A Postgres instance. Temper itself, serving its CLI, API, and MCP. Dave provisioned, org-common created, a webhook wired, and a threshold that woke onboarding-agent#1 for the first time. None of that appears in the story the show pages told, and all of it had to happen for the story to be possible. And "something had to be running" hides a choice: running how, and where?

That's the turn outward — from a map whose shape is proven to the good problem of standing one up and keeping it running. It's a different reader's question, too: not the journeyer who came to understand a cognitive map, but the evaluator who arrives cold to ask what they'd run, what they get out of the box, and what stays theirs to decide.

Walked through under /operating

So the operating story has its own home now — told from that cold start rather than carried as the tail of this journey. Operating Temper sorts it against a three-tier confidence ledger — what the architecture fixes, what's extensible with the edges still being found, and what's genuinely yours — and splits four ways:

  • Deployment — topology, tenancy, per-tenant integration, and where agents run.
  • Governance & administration — authoring versus reshaping the access graph, and how guarded that must be.
  • Observability & audit — is it healthy, and what did it know and why did it change.
  • Insights — the queryable provenance of how a shared understanding formed. The forward-looking close.

That's the invitation, in the end: not only to help build a system that grows understanding, but to help run it where the operating shape is yours to choose.