Overview

Route inference and profiling campaign information to their domain owners; keep Memory pointers and reusable lessons. Insight already contains the pilot record, so no destination migration or feed change is necessary.

Plan

1. Brain ORGANISM.md: explicit routing by information type, not permanence, preserving Cortex science records and Mind implementation history.
2. Brain skills/sampler_pipeline/{SKILL.md,reference.md}: replace blanket durable-findings-to-Memory instructions with Insight/inference and Pulse/profiling routes; Memory only reusable lessons and pointers.
3. Memory AGENTS.md: matching write boundary. Replace only the A5 catalogue summary in wiki/methods/concepts/sampler-benchmarks.md with links to Insight wave-1 record and project-owned catalogue/protocol; update index wording and append log provenance. Preserve older historical evidence pending separate provenance migration.
4. Validate Memory with make validate, Brain documentation/skill checks applicable to changed files, verify destination links from GitHub and review the diff. No scientific API changes; downstream smoke n/a.

Branch survey

Brain: main, clean. Memory: main, unrelated bibliography and lensing edits; use isolated worktrees. Insight and Pulse: main, clean but claimed by search-ext-b5b2-remainder and ral-category-paths respectively; read-only destinations, no claims requested.

Branch: feature/memory-domain-routing
Worktree: .worktrees/memory-domain-routing (inside workspace)
Tier: judge — merge mode: human /prm

Heart RED development override

After the exact reasons were presented, the live human answered "Proceed" to development and PR opening for this task. No merge or release; applicable checks must pass. Validation pending implementation.

  [red] release validation FAILED (stage integrate)
  [red] PyAutoHands/workflows/Python Version Matrix: failure
  [red] PyAutoHeart/workflows/Release Integrate: failure
  [red] PyAutoMind/workflows/Lifecycle Drift: failure
  [red] PyAutoArray/Tests: red
  [red] PyAutoArray/Tests: red
  [red] autofit_workspace_test/Smoke Tests: red
  [red] autogalaxy_workspace_test/Smoke Tests: red
  [red] autolens_workspace_test/Smoke Tests: red
  [red] CI wall-clock: 26 gates · slowest PyAutoBrain Nightly Release 63m · 5 hang events
  [red] Release readiness: release validation FAILED (stage integrate)
  [red] autofit_test: red
  [red] Release validation: NOT release_ready — v2026.10.15.1.dev81701 profile=release (2026-10-15T07:04:34+00:00)
  [red] validation_report/stages/integrate: fail
  [red] PyAutoArray: red
  [red] autolens_inference: red
  [red] autolens_profiling: red
  [red] Worktree drift: 5 orphan / 0 missing / 4 dirty
  [red] euclid_strong_lens_modeling_pipeline: red

Original prompt and accepted scope

Route campaign information to Insight and Pulse

Type: docs
Target: brain
Difficulty: medium
Consequence: judge
Autonomy: supervised

Original request (verbatim)

The following appears on the live feed of pyautolabs / science, but I think the work its describing belongs under PyAutoInsight, so I wonder if we need to work on routing of memory / permanant information to the right organ in certain cases. I suspect a similar thing could happen with profiling / speed up of code not being routed to Pulse? 📚 Memory
docs(methods): sampler-benchmarks points at the autofit inference search catalogue (search-ext A5 consumers)
PR opened
3h ago
PyAutoLabs/PyAutoMemory · open

3h ago PR opened

Accepted scope

User agreed that project-owned detailed results stay in autofit_inference, Insight owns campaign records and evidence presentation, and Memory retains pointers plus reusable scientific lessons. User: "ok great proceeed". After the exact Heart RED reasons were reported, user: "Proceed" (development and PR opening only; no merge).

Change @PyAutoBrain routing guidance and @PyAutoMemory's duplicate A5 catalogue section. Insight already has the pilot record in tasks/gaussian_x3_search_wave1.md on main, including 223/520 and no-ranking interpretation; retain that destination and link it rather than creating another summary. Pulse owns profiling campaign intent and evidence presentation; project profiling repos retain measurements; library performance implementation stays in its code repo. No feed changes, compute, historical bulk migration or edits to claimed Insight/Pulse repos.

Implementation plan

1. Brain ORGANISM.md: explicit routing by information type, not permanence, preserving Cortex science records and Mind implementation history.
2. Brain skills/sampler_pipeline/{SKILL.md,reference.md}: replace blanket durable-findings-to-Memory instructions with Insight/inference and Pulse/profiling routes; Memory only reusable lessons and pointers.
3. Memory AGENTS.md: matching write boundary. Replace only the A5 catalogue summary in wiki/methods/concepts/sampler-benchmarks.md with links to Insight wave-1 record and project-owned catalogue/protocol; update index wording and append log provenance. Preserve older historical evidence pending separate provenance migration.
4. Validate Memory with make validate, Brain documentation/skill checks applicable to changed files, verify destination links from GitHub and review the diff. No scientific API changes; downstream smoke n/a.

Branch survey

Brain: main, clean. Memory: main, unrelated bibliography and lensing edits; use isolated worktrees. Insight and Pulse: main, clean but claimed by search-ext-b5b2-remainder and ral-category-paths respectively; read-only destinations, no claims requested.

Branch: feature/memory-domain-routing
Worktree: .worktrees/memory-domain-routing (inside workspace)
Tier: judge — merge mode: human /prm

Test Plan

• Brain full suite: 1,264 passed; one environment-dependent failure in test_hygiene_map_and_heart_board_find_grouped_checkouts because task activation exports PYAUTO_MIND, overriding its temporary fixture.
• Corrected invocation: env -u PYAUTO_MIND python -m pytest tests/test_grouped_organ_consumers.py -q — all 4 passed. No source/test changes or skipped tests. All collected cases have passing evidence across these runs.
• Memory: 276 tests passed; make validate passed (citations, structure, no new unresolved wikilinks).
• Brain tenant firewall and CI-scoped discovery checks (Brain + Mind) passed. Broader discovery scan reports pre-existing Memory missing catch-up links and autofit_assistant drift outside this diff.
• All five new GitHub destination links resolve; pinned catalogue confirms the removed counts/statuses remain preserved at source.
• Author diff review: no findings; not an independent review (ordinary human-approved docs task). Both diffs pass git diff --check.
• Local evidence: .worktrees/memory-domain-routing/logs/ (brain-tests.log, memory-tests.log, memory-validate.log, firewall.log, review.md, readiness.json).

Delivered

• Brain PR: docs: route campaign records through Insight and Pulse #525 (bc3843c)
• Memory PR: docs: point sampler campaign records to Insight PyAutoMemory#133 (165be2f)
• Both remain open for human /prm; no merge/release authorized.