[Jammy2211]

Summary

Shadow shift 1 of the pilot (#574) produced a digest that couldn't be reviewed. It listed 6 bare prompt slugs, each with why: scoped tier notify, and nine zero counters. The shadow pilot row also measured the wrong thing. This PR:

• names each pick by its title, linked to its prompt, with size and repos;
• adds the planner's real ranking to why: (queue.md position, or "not in queue.md" with review-minutes and priority, plus epic phase);
• collapses a shadow digest's counters to one line;
• leaves a shadow pilot row's merges as ?, so the human records their own merges in the window (the epic's baseline: "the human works as today, minutes are logged").

Closes #579

API Changes

Adds _orchestrator.pick_record, _brief.rank_note, _digest.label, _digest.mind_blob_base and _digest.SHADOW_COUNTERS. _digest.build gains an issue= argument. survey records now carry title, and planner records carry queued. pick_reason lines are longer (rule · ranking). Older digests with untitled picks still validate.

Witness

The shift-2 brief rendered on the real backlog (a shadow tick into a scratch state folder; Mind untouched):

## Digest

### Brief work
- [autocti_workspace_test: pin numerical log-likelihood values for the CTI likelihood functions](https://github.com/PyAutoLabs/PyAutoMind/blob/main/draft/test/autocti_workspace_test/likelihood_value_pins.md) (`likelihood_value_pins`; small · autocti, autocti_workspace_test) — proposed (shadow) — glance
- [`imaging/features/pixelization/delaunay.py` sits on the 300 s smoke cap (330–354 s under the release profile)](https://github.com/PyAutoLabs/PyAutoMind/blob/main/draft/bug/autolens_workspace/delaunay_smoke_300s_cap.md) (`delaunay_smoke_300s_cap`; small · autolens_workspace) — proposed (shadow) — glance

### Backlog picks
- none

### Proof packets
- none — shadow shift: nothing was dispatched

### Counters
- all counters 0 — shadow shift, nothing dispatched · review-minutes-planned: 20

Shift 1's unguided picks, re-rendered in the new format. The why: now shows they all tie at ~0 review-min with no queue.md entry, so the order fell to the file path (alphabetical):

- [`batch slice` — the decomposition pass for `needs-slicing` prompts](https://github.com/PyAutoLabs/PyAutoMind/blob/main/draft/feature/pyautobrain/batch_slice.md) (`batch_slice`; medium · pyautobrain, pyautomind) — proposed (shadow) — notify — why: scoped tier notify · not in queue.md · ~0 review-min · priority normal
- [Board phase 2: the remaining four legs onto the seam](https://github.com/PyAutoLabs/PyAutoMind/blob/main/draft/feature/pyautobrain/board_without_gh_phase2_legs.md) (`board_without_gh_phase2_legs`; small · pyautobrain) — proposed (shadow) — notify — why: scoped tier notify · not in queue.md · ~0 review-min · priority normal
- [autofit_assistant: a skill that owns loading the user's data and its selection cuts](https://github.com/PyAutoLabs/PyAutoMind/blob/main/draft/feature/autofit_assistant/data_loading_and_selection_cuts_skill.md) (`data_loading_and_selection_cuts_skill`; small · autofit_assistant) — proposed (shadow) — notify — why: scoped tier notify · not in queue.md · ~0 review-min · priority medium
- [Demo `subplot_ellipse_errors` from a real ellipse model-fit](https://github.com/PyAutoLabs/PyAutoMind/blob/main/draft/docs/workspaces/demo_subplot_ellipse_errors.md) (`demo_subplot_ellipse_errors`; large · autogalaxy_workspace) — proposed (shadow) — notify — why: scoped tier notify · not in queue.md · ~0 review-min · priority normal
- [Regenerate autolens_workspace markdown/ so the MGE pages show sigma_min](https://github.com/PyAutoLabs/PyAutoMind/blob/main/draft/docs/autolens_workspace/markdown_regeneration_sigma_min.md) (`markdown_regeneration_sigma_min`; small · autolens_workspace) — proposed (shadow) — notify — why: scoped tier notify · not in queue.md · ~0 review-min · priority normal
- [RTD organism docs currency: Nerves page, organ-count drift, hands.md rename](https://github.com/PyAutoLabs/PyAutoMind/blob/main/draft/docs/pyautobrain/rtd_organism_currency.md) (`rtd_organism_currency`; small · pyautobrain) — proposed (shadow) — notify — why: scoped tier notify · not in queue.md · ~0 review-min · priority normal

Tests

• Updated digest/shadow tests to the titled format.
• New tests: titled picks validate; a missing why: names the slug; the one-line shadow counters are accepted only as that line; an old untitled digest still validates; a live pilot row counts merges while a shadow row leaves ?; batch watch shows titles.
• Full Brain suite: 1564 passed, 1 skipped. Tenant firewall: OK. The real shift-1 record still passes batch brief --check.
• Heart: GREEN (release).

Tier: glance — merge mode: auto-merge on green with Witness PASS, post-merge summary.

🤖 Generated with Claude Code