nx_maint_research_fetch.nx

nx_maint_research_fetch.nx -- SOVEREIGN multi-source web-fetch to RESEARCH how S-class systems CLOSE THE FEEDBACK LOOP on physical things (wear/design/ capability), so the maintenance platform can be SIMULATED + proven in software (no IRL spend) and EXCEED open-loop tools like OBD (which only throw errors and need a chain of 3rd parties to be meaningful/actionable/reproducible). Operator (2026-06-23): "use nishi researcher to see how s class close the feedback loops on physical things ... full learning loop ... simulate ... till proven via modeling." Mirrors nx_iot_research_fetch: own TLS-1.3 + Mozilla CA store, saves each source to knowledge/fetched/maint_*.raw (research COMPOUNDS, >=2-source cite). 100% sovereign (no curl/wget/gcc). Idempotent have-skip. Axes = what a CLOSED-LOOP, SIMULATE-FIRST maintenance platform must learn: (A) DIGITAL TWIN / SIMULATION (model the physical asset in software) (B) FEEDBACK / CLOSED-LOOP CONTROL (the loop itself: observe->act->observe) (C) PREDICTIVE MAINTENANCE / PROGNOSTICS (the maintenance learning loop, RUL) (D) OPEN-LOOP DIAGNOSTICS (OBD) (what we EXCEED -- errors w/o a closed loop) (E) LEARNING / WEAR MODELING (system-id, state estimation, failure rate) expect_exit: 0 license_tier: ORIGINAL