Essays, use cases, and notes on capturing agent work and progressively formalizing reliable processes.https://functionary.app/https://functionary.app/blog/safety-is-a-road/https://functionary.app/blog/safety-is-a-road/Legible structure helps agents stay on course. Capability roots, Test-mode stubbing, and Production approval also enforce boundaries they cannot cross.Mon, 01 Jun 2026 00:00:00 GMTessayhttps://functionary.app/blog/judgment-is-not-delegable/https://functionary.app/blog/judgment-is-not-delegable/Accountability cannot be transferred to a model. A tool for serious work has to be built around that fact.Sun, 24 May 2026 00:00:00 GMTessayhttps://functionary.app/blog/analyst-weekly-memo/https://functionary.app/blog/analyst-weekly-memo/A linear pipeline that turns a week of source material into a reviewed, sent memo — with a human sign-off in the middle.Fri, 22 May 2026 00:00:00 GMTusecasehttps://functionary.app/blog/intake-triage/https://functionary.app/blog/intake-triage/A branching pipeline that classifies incoming requests, routes by urgency, and escalates the hard cases to a person.Mon, 18 May 2026 00:00:00 GMTusecasehttps://functionary.app/blog/at-least-once-over-exactly-once/https://functionary.app/blog/at-least-once-over-exactly-once/Pure work can retry freely. External writes need an explicit replay protocol, because a crash can leave the truth unknowable.Sat, 16 May 2026 00:00:00 GMTnotehttps://functionary.app/blog/feedback-synthesis/https://functionary.app/blog/feedback-synthesis/A fan-in pipeline that merges several feedback streams, extracts entities, compares across sources, and validates before it writes.Tue, 12 May 2026 00:00:00 GMTusecasehttps://functionary.app/blog/verifiability-is-not-trust/https://functionary.app/blog/verifiability-is-not-trust/Being able to check an output after the fact is not the same as being able to trust the process that produced it.Fri, 08 May 2026 00:00:00 GMTessay