Include the current addon, terrain integrations, tools, regression tests and recorded development evidence.
16 lines
1.5 KiB
JSON
16 lines
1.5 KiB
JSON
{
|
|
"scope": "Ordinary ground memory SEARCH only: snapshot finite delivered-belief uncertainty clamped45..110m at Assign/Retask, validate copied radius before spawn, set once after successful spawn and before old assignment release/new installation. No live read of player location, new actor, settings, radius animation or deadline extension.",
|
|
"exclusions": [
|
|
"Missing-patrol investigation and road/site-security Defend tasks retain existing defaults.",
|
|
"INTERCEPT/CONTAIN and air direct-task path remain unchanged.",
|
|
"WORKBENCH NativePatrolHQProbe now initializes the same snapshot for its direct ordinary pending-task fixture; no native run."
|
|
],
|
|
"proof": "Six focused programs green; old HQ actual DeliverOrder compiles and runtime-fails radius assertion. Omitted finite/range admission and omitted Retask snapshot each compile/runtime-fail. Existing actual40m/30s negative-sweep witness remains byte-identical and tested separately; larger native search grid does not prove larger cleared area.",
|
|
"validation_limits": [
|
|
"Native AI/waypoint/group boundaries are doubles; parser/radius policy and actual HQ delivery/queue expressions run as C++.",
|
|
"Grid grows with radius;110m is a size bound, not a measured performance bound or FPS improvement.",
|
|
"Native compilation, actual search movement and multiplayer remain unqualified.",
|
|
"Historical prepare.py/adapt-tests.py were construction scratch; only manifest-listed files are authoritative."
|
|
]
|
|
}
|