Skip to content

docs(r2-structure): allow pre-R1-close spawn for 5 of 6 R2 managers (PB stays gated) - #873

Merged
briansrls merged 1 commit into
mainfrom
pm/r2-structure-pre-r1-spawn-amendment
Apr 26, 2026
Merged

briansrls merged 1 commit into
mainfrom
pm/r2-structure-pre-r1-spawn-amendment

Conversation

@briansrls

Copy link
Copy Markdown
Contributor

Summary

Per-manager R1-gating audit (2026-04-26) found only Pure Bootstrap is definitionally R1-close-gated — its scope is "what survives R1 close" (mirror dissolutions, Tier 2 patch retirement, post-R1 emergent dissolutions). The other 5 R2 managers (Grounding, Substrate, Modeling, Impossible-Bugs, Release) have no technical R1 dependency.

In particular, R2 Substrate Manager's ValueBody-list/sum sub-lane is a PREREQ for R1C-A Sub-deliverable A (M1(2.8) list-body lowering depends on the substrate variant landing) — pre-R1-close Substrate spawn actively unblocks R1 closure rather than competing with it.

This amendment refines r2-structure.md Transition mechanics step 4's framing to allow pre-R1-close spawn for the 5 non-PB managers, while preserving the load-bearing dissolution discipline (step 3 R1 manager dissolution still happens at R1 gates green; step 5 ROADMAP promotion still gates on R1 close).

Surface edits (3 total)

  1. Transition mechanics step 4 — appended 🔄 REVISED 2026-04-26 paragraph stating PB is the only definitionally R1-close-gated manager; other 5 may spawn pre-R1-close when (a) brief queue authored, (b) at least one deliverable dispatchable, (c) no conflict with R1 closure.
  2. Transition mechanics step 5 — added clarification that ROADMAP ## Release R2 Program promotion still gates on R1 close; pre-R1-close R2 manager activity operates under this doc's PROPOSAL authority.
  3. Decisions locked — added **Pre-R1-close R2 manager spawn allowed for 5 of 6 managers** entry. Documents audit + locked decision + sequencing detail (PROPOSAL vs ROADMAP authority phases).

Why now

Tonight (2026-04-26 ~06:30Z), PM staffed R1 Closure Manager (fierce-raven-707 #849) and briefed Director (zesty-bear-812 #828) to spawn the 5 R2 managers in parallel with R1 closure work. Amendment lands the doc-level authority for that staffing decision so practice + authority don't drift.

Per-manager R1-gating receipts

R2 Manager R1-gated? Why
Grounding NO Already partially in flight (Pilot PR #765, Engine Phase 1 PR #788).
Substrate NO + actively unblocks R1 ValueBody-list/sum is prereq for R1C-A Sub-deliverable A.
Modeling NO Substrate-gated per item. Dimensions ungated by audit (Director reframe #4).
Impossible-Bugs NO All 3 PR #836 implementation workers explicitly UNGATED per design-doc audits.
Release Mostly NO §6a follow-through, B5/B6/B7 dispatchable. Thesis-claim mapping IS R1-close-gated (Open call 1).
Pure Bootstrap YES Definitionally — scope is "what survives R1 close."

Out of scope (intentional non-edits)

  • Manager structure section — 6-manager structure is unchanged.
  • PB Manager scope — still post-R1 only (load-bearing definitional anchor).
  • Demo discipline / closure ledger / velocity tripwire — Release Manager standing duties unchanged whether spawn is pre- or post-R1-close.

Per-PR dissolution gate

  • Defers / shrinks / deletes: N/A — docs-only; surface change minimal (1 paragraph append + 1 clause append + 1 Decisions-locked entry). 4 net additions across 1 file.

Test plan

  • Verify the amendment paragraph in step 4 doesn't supersede step 3's R1 manager dissolution discipline.
  • Verify the amendment doesn't change PB Manager's scope (still post-R1 only).
  • Verify the new Decisions-locked entry references the audit + sequencing correctly.

🤖 Generated with Claude Code

…PB stays gated)

Per-manager R1-gating audit (2026-04-26) found only Pure Bootstrap
Manager is definitionally R1-close-gated — its scope is "what survives
R1 close" (mirror dissolutions, Tier 2 patch retirement, post-R1
emergent dissolutions); R1's PB gates (T-PB-A + T-PB-B census reduction)
own R1-side work, and PB Manager picks up the residual.

The other 5 managers (Grounding, Substrate, Modeling, Impossible-Bugs,
Release) have no technical R1 dependency. R2 Substrate spawn actively
*unblocks* R1 closure: T-Substrate ValueBody-list/sum is a PREREQ for
R1C-A Sub-deliverable A (M1(2.8) list-body lowering depends on the
substrate variant landing).

## Two surface edits

1. Transition mechanics step 4 — appended a "🔄 REVISED 2026-04-26"
   paragraph stating PB is the only definitionally R1-close-gated
   manager; other 5 may spawn pre-R1-close when (a) brief queue is
   authored, (b) at least one owned deliverable is dispatchable,
   (c) spawning doesn't conflict with R1 closure work. Step 3's R1
   manager dissolution discipline is preserved (still happens at R1
   gates green).

2. Transition mechanics step 5 — added clarification that ROADMAP
   `## Release R2 Program` promotion still gates on R1 close;
   pre-R1-close R2 manager activity operates under this doc's PROPOSAL
   authority.

3. Decisions locked — added "Pre-R1-close R2 manager spawn allowed
   for 5 of 6 managers" entry. Documents the audit + locked decision +
   sequencing detail (PROPOSAL vs ROADMAP authority phases).

## Why now

Tonight (2026-04-26 ~06:30Z), PM staffed R1 Closure Manager
(`fierce-raven-707` #849) and briefed Director (`zesty-bear-812` #828)
to spawn the 5 R2 managers in parallel with R1 closure work. The
amendment lands the doc-level authority for that staffing decision so
practice and authority don't drift. Director's brief explicitly cites
this amendment as the authority basis for the staffing call.

## Out of scope

Scope-narrow surface edit; intentionally NOT touching:
- Manager structure section (the 6-manager structure is unchanged)
- PB Manager's scope (still post-R1 only — that's the load-bearing
  definitional anchor)
- Demo discipline / closure ledger / velocity tripwire (Release
  Manager standing duties are unchanged whether spawn is pre- or
  post-R1-close)

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@briansrls

Copy link
Copy Markdown
Contributor Author

Review metadata

  • Provider / model: openai-pro / gpt-5-5-pro
  • Commit: d3cb823e · Trigger: manual
  • Conversation: View conversation

1. Story of the diff

This PR changes docs/r2-structure.md so the R1→R2 transition is no longer “all six R2 managers wait until R1 close.” The new contract keeps Pure Bootstrap gated on R1 close because its scope is explicitly residual—“what survives R1 close”—but allows the other five managers to spawn before R1 closes when their briefs are authored, at least one deliverable is dispatchable, and the work does not compete with R1 closure (docs/r2-structure.md:257). It also records that R2 Substrate may actively unblock R1 because the ValueBody list/sum lane is a prerequisite for R1C-A Sub-deliverable A (docs/r2-structure.md:257). The PR preserves the release-authority boundary by keeping ROADMAP promotion gated on R1 close and stating that pre-close activity remains under this proposal doc’s authority (docs/r2-structure.md:258, docs/r2-structure.md:286).

2. Invariant categories

  1. LAYER MODEL (substrate vs implementation).

N/A — diff is documentation-only in docs/r2-structure.md; it does not touch Dag, substrate carriers, cross-pass data, or implementation Rust.

  1. INVARIANTS.md + modeling-discipline.md.

Compliant — P5 Progress Is Dissolution / Scaffold Boundaries are preserved: the pre-R1-close exception is bounded to five named managers, requires authored briefs + dispatchable work + no R1 conflict, and explicitly says it does not supersede R1 manager dissolution (docs/r2-structure.md:257). P2 Boundary Discipline is also preserved by keeping ROADMAP authority gated on R1 close while pre-close work runs only under proposal authority (docs/r2-structure.md:258, docs/r2-structure.md:286).

  1. CODING.md.

N/A — no Rust implementation, helper placement, method/free-function shape, error carrier, or naming surface is changed.

  1. TESTING.md.

N/A — no executable compiler behavior changes. A doc-only sequencing clarification does not require unit, integration, or boundary tests.

  1. LOCKED DESIGN DECISIONS.

Compliant — the divergence from the prior “spawn at R1 close” framing is explicit and dated as locked user direction (docs/r2-structure.md:286), while the ROADMAP promotion gate remains unchanged (docs/r2-structure.md:258).

  1. TRACKED vs UNTRACKED DEBT.

Compliant — the temporary/pre-close operating mode is tracked rather than open-ended: bounded to 5 of 6 managers and three spawn predicates (docs/r2-structure.md:257), with authority bounded to proposal status until the R1-close ROADMAP promotion point (docs/r2-structure.md:258, docs/r2-structure.md:286). No new TODOs, unnamed scaffolds, or unbounded bridges are introduced.

3. Verdict

APPROVE

The PR is a narrow documentation correction that makes the manager-spawn exception explicit, bounded, and authority-scoped. I do not see a substrate, testing, coding-style, or untracked-debt issue in the diff.

@briansrls briansrls left a comment

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review metadata

  • Provider / model: codex / unknown
  • Commit: d3cb823e · Trigger: schedule
  • Thinking: 134s wall

Non-blocking — Strengths

  • docs/r2-structure.md Classified as design/docs; the amendment preserves R1 closure, R1 manager dissolution, and ROADMAP promotion gates while allowing bounded pre-spawn under PROPOSAL authority.

Non-blocking — Improvements (fix in-PR if easy, else defer to roadmap)

  • docs/r2-structure.md The new locked decision says the other five managers have no technical R1 dependency, but Release still has thesis-claim mapping tied to R1 closure; narrow this to manager-spawn eligibility if deferred (roadmap lane: R2 promotion hygiene).

ROADMAP — Verified

  • R2 promotion gate: The added transition text keeps ROADMAP Release R2 Program promotion gated on R1 close while allowing pre-R1 work only under this proposal document.

✅ No blocking concerns.

@briansrls
briansrls merged commit 976b1e2 into main Apr 26, 2026
4 checks passed
@briansrls
briansrls deleted the pm/r2-structure-pre-r1-spawn-amendment branch June 1, 2026 18:41
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant