docs: record raw pilot implementation and development checks

campaigns.yaml

@@ -122,10 +122,11 @@ tasks:
   priority: high
   path: tasks/hst_mge_cpu_pilot.md
   updated: '2026-10-13'
-  next: Review the draft pilot, resolve host/stack/data/witness choices and plan the bounded project producer.
-    Proposed maximum 32 sequential processes, 16 CPU-hours; compute and freeze remain unauthorized.
-  issue: https://github.com/PyAutoLabs/PyAutoPulse/issues/53
-  evidence: https://github.com/PyAutoLabs/autolens_profiling/pull/414
+  next: Preparation merged (#54); raw fixed-parameter runner implementation approved and tracked in
+    autolens_profiling#415. Review producer evidence, then resolve exact host/stack/data/witness choices.
+    No inference, production compute, baseline freeze or scientific acceptance authorized.
+  issue: https://github.com/PyAutoLabs/autolens_profiling/issues/415
+  evidence: https://github.com/PyAutoLabs/PyAutoPulse/pull/54
 - id: runtime_cell_single_jit_gpu_warmup
   campaign: point-source-source
   title: Runtime cells' A100 `single_jit` includes the post-compile warm-up — source-plane 0.642 ms vs a steady

dashboard.md

@@ -1,6 +1,6 @@
 # PyAutoPulse — profiling dashboard
 
-<!-- pulse-campaigns:b22a787746f3dc695099e748c7d1bb4ca52c932cb99185bf625878506ca73f04 -->
+<!-- pulse-campaigns:087d07eeb8a1b75b8ed430161548d93c3d73d8be8d9b8db6981565b5f86ddcc0 -->
 ## Profiling Check In
 
 <details><summary>Full check-in prompt</summary>
@@ -38,7 +38,7 @@ Reviewed 2026-10-13. HST/MGE fp64 CPU pilot preparation approved; review its dra
 ### Active tasks
 
 - [Freeze the setup baseline specification, then collect and review evidence](https://github.com/PyAutoLabs/PyAutoPulse/blob/main/tasks/setup_baseline_campaign.md) — needs-decision: Prepare the HST/MGE CPU pilot first; review its protocol/resource proposal and resolve exact host/stack/data/witness choices. Manual HST builder shipped, but collection producer remains pending. Wider specification stays draft; collection, scientific acceptance and baseline promotion are separate.
-- [Prepare and review the HST/MGE fp64 CPU pilot](https://github.com/PyAutoLabs/PyAutoPulse/blob/main/tasks/hst_mge_cpu_pilot.md) — needs-decision: Review the draft pilot, resolve host/stack/data/witness choices and plan the bounded project producer. Proposed maximum 32 sequential processes, 16 CPU-hours; compute and freeze remain unauthorized.
+- [Prepare and review the HST/MGE fp64 CPU pilot](https://github.com/PyAutoLabs/PyAutoPulse/blob/main/tasks/hst_mge_cpu_pilot.md) — needs-decision: Preparation merged (#54); raw fixed-parameter runner implementation approved and tracked in autolens_profiling#415. Review producer evidence, then resolve exact host/stack/data/witness choices. No inference, production compute, baseline freeze or scientific acceptance authorized.
 
 </details>
 

tasks/hst_mge_cpu_pilot.md

@@ -157,3 +157,35 @@ Reviewed 2026-10-13 at project merge `b8b3b23`; source review does not refresh m
 - [Baseline procedure](https://github.com/PyAutoLabs/autolens_profiling/blob/b8b3b23cd1f0aabc2d7d4ce1cd297f7420b14f76/baseline/README.md)
 - [Draft baseline manifest](https://github.com/PyAutoLabs/autolens_profiling/blob/b8b3b23cd1f0aabc2d7d4ce1cd297f7420b14f76/baseline/campaign.json)
 - [Global route registry](https://github.com/PyAutoLabs/autolens_profiling/blob/b8b3b23cd1f0aabc2d7d4ce1cd297f7420b14f76/catalogue/script_routes.json)
+
+
+## Implementation handoff — 2026-10-13
+
+Preparation [Pulse #54](https://github.com/PyAutoLabs/PyAutoPulse/pull/54) merged
+at `71c93fc285a081ffb1f0f506694c043a15ad4e22`; bounded implementation approved by
+the human ("ok continue"). [Project #415](https://github.com/PyAutoLabs/autolens_profiling/issues/415)
+owns a separate raw pilot runner: fixed parameters, fresh processes, unrounded
+phase timings, target-specific cache diagnostics and actual-context checks.
+No inference is performed. The runner's records will remain `accepted: false`;
+independent scientific correctness and validated background-load methodology
+remain prerequisites for collection qualification, not inferred implementation
+successes. The full baseline reader/manifest stays unchanged.
+
+The producer moved to [Mind active work](https://github.com/PyAutoLabs/PyAutoMind/blob/main/active/hst_mge_pilot_observations.md).
+Production compute, campaign freeze and scientific acceptance remain unauthorized.
+
+
+### Development evidence — 2026-10-13
+
+At project commit `b5b8160b23d9a59bb9cb712c1d7bdd35198fbad1`, a local CPU development
+fixture (11x11 synthetic image, 3x3 PSF, fp64, one repetition, 60-second per-worker
+cap) completed cold, seed, warm and runtime in four distinct processes. The warm
+worker reported the target executable's cache hit with the seed's matching key;
+the runtime worker retained ten individual synchronized samples. All records
+remained `accepted: false`. This verifies producer mechanics only: it supplies no
+production timing comparison, independent scientific witness or accepted baseline.
+The 32 protocol tests passed and independent code review was CLEAN after rejecting
+dirty/unknown source identities. Project #415 retains the implementation and PR
+validation record; full scientific qualification remains pending here.
+
+Implementation PR: [autolens_profiling #416](https://github.com/PyAutoLabs/autolens_profiling/pull/416), opened 2026-10-13; pending CI/merge.