[Goal] Derive a parameter-closed emergent-gravity stack from accepted substrate dynamics

[vantasnerdan]

Positive objective

Build and promote the framework's first parameter-closed emergent-gravity stack: start from accepted substrate dynamics, derive rather than assume a generally covariant low-energy gravitational action and its normalization, obtain a controlled nonflat sourced spacetime, and show that both massive and massless probes follow from the same derived geometry. The completed construction must make at least one static and one radiative held-out prediction from the same frozen parameter ledger.

This is an objective-level successor to the unresolved frontier documented in #12 and must respect the Route-2 no-go established in #9. It may reuse their correct artifacts and rejected-route evidence, but neither issue's narrative is authority.

Why this is the highest-value next goal

Release v0.160.0 now contains strong but separate pieces:

• conditional induced-coupling and identifiability bookkeeping (C-GRV-001);
• exact effective geometry without physical gravity (C-GOR-001);
• Einstein-scalar consumers whose gravitational coupling is supplied (C-STG-001, C-STG-002);
• conditional static/radiative geometry and quadrupole consumers (C-GW-*);
• accepted massive, massless, reparametrization, and conformal worldline mechanics (C-WLN-001..003).

The framework still lacks the load-bearing bridge that makes these one theory: no accepted substrate mechanism derives the gravitational action, fixes its usable normalization and corrections, generates a nontrivial physical spacetime, and sends both matter and radiation through that same result. Closing that bridge would turn several conditional sectors into a coherent predictive stack.

What counts as real success

The issue closes only when one coherent, independently reviewed construction does all of the following:

1. Derives a low-energy metric/connection action from accepted framework degrees of freedom and explicitly approved imports. Its leading regime is generally covariant and Einstein-equivalent, or a separately governed foundational revision proves that the smallest coherent alternative is required.
2. Derives the usable gravitational normalization and the regime of validity. A free additive baseline, fitted Newton constant, borrowed Einstein-Hilbert coefficient, arbitrary field-to-metric rescaling, or comparator-selected regulator/field content does not close the goal. Any renormalization condition must itself have declared, reviewable provenance.
3. Controls every term that can defeat the claimed regime, including the cosmological and leading higher-curvature/dispersion/Lorentz-violating sectors as applicable. They may be derived, cancelled by a justified mechanism, or bounded in a predeclared domain; they may not be omitted silently.
4. Produces at least one controlled, nonflat sourced solution of the derived field equations, with conservation, boundary/initial data, existence or numerical-convergence evidence, and known-limit checks appropriate to the claim.
5. Composes C-WLN-001..003 with that same solution so massive and massless probes obtain the correct timelike/null dynamics without supplying a second geometry or force normalization.
6. Produces at least two independent consequences from the same frozen inputs: one static/weak-field observable and one radiative/dynamical observable. At least one must be confronted with a held-out empirical or independently established GR benchmark under a preregistered error model. Algebraically equivalent restatements do not count as independent predictions.
7. Promotes only the claims that individually pass the repository's success definition: importable APIs and tests, sensitive strongest-practical oracles, independent rederivation, full dependency replay, claim-level review, synchronized registry/release/generated docs/memory, and no debt inside the promoted scope.

A no-go, wrong sign, uncontrolled correction, nonidentifiability result, failed comparator, or useful partial API is durable attempt/harvest evidence. It advances the campaign but is not completion; continue with a materially different candidate unless the user changes the objective.

Scientific freedom

The implementing agent owns the scientific strategy. This issue does not prescribe an induced-action route, regulator, field content, ansatz, discretization, representation, or proposal sequence. Compare plausible concepts under the candidate-first workflow and let framework fit and evidence select the route. Do not retrofit accepted invariants around a favored mechanism.

Accepted base is v0.160.0. Audit before reuse at minimum:

• governance/releases/current.yaml and governance/claims.yaml;
• [Campaign 2] Build the 1+1 → 3+1 sine-Gordon → collective-tensor bridge toward GR #9 and [Campaign 3] Induced (Sakharov) gravity from the sine-Gordon substrate — derive 1/G from the fluctuation spectrum #12, including their harvest handoffs and counterexamples;
• src/substrate_framework/covariant_sine_gordon_action.py and scalar_induced_newton.py as unpromoted/review-bounded prior work;
• the canonical modules behind C-GRV-001, C-GOR-001, C-STG-001/002, C-GW-*, and C-WLN-001..003.

These are starting evidence, not mandatory ingredients. Reuse only what survives source audit and exact dependency matching.

Custom IPython harness contract

The stress-test harness must expose the entire repository workflow catalog to the agent and record which routes were actually invoked. Native skill discovery is preferred; otherwise it must resolve and read the corresponding SKILL.md in full. Reference skill names rather than hard-coding this workstation's absolute paths.

Required instruction and skill references

• AGENTS.md — normative scientific, governance, validation, and non-self-merge contract.
• AGENTS_START_HERE.md — authority, issue/branch coordination, memory, GitNexus, PR, review, and handoff workflow.
• physics-erdos-loop — .agents/skills/physics-erdos-loop/SKILL.md; mandatory for this campaign, including every routed reference needed by the selected oracle and phase.
• research-pr-harvest — .agents/skills/research-pr-harvest/SKILL.md; mandatory when a scientific PR is reviewed, harvested, narrowed, or handed off.
• shared_agent-memory-usage — tools/agent-memory/skills/shared_agent-memory-usage/SKILL.md when the harness needs memory CLI syntax; repository memory remains an index/work record, not authority.
• GitNexus skill catalog: gitnexus-guide, gitnexus-cli, gitnexus-exploring, gitnexus-impact-analysis, gitnexus-debugging, gitnexus-refactoring, and gitnexus-pr-review. The harness must advertise all seven and invoke each when its trigger applies; merely listing them is not evidence that their workflow ran.

Required workflow behavior

1. Clean startup: show repository root, exact commit/branch, dirty files, current release, open issues/PRs, resolved instruction files, available tools, and resolved skill catalog before substantive work.
2. Authority and recall: read accepted release/claims and relevant canonical sources; search repository memory and verify useful hits at source; inspect both source text and the refreshed GitNexus graph.
3. Durable contracts before computation: instantiate the research-arc, effort, and campaign proposal contracts that apply; validate the proposal manifest/schema before opening source bodies or comparator values. Register competing candidates and structural criteria first.
4. Blinding: the harness must be able to withhold empirical comparator values until the proposal's equations, conventions, candidates, and selection criteria are frozen. Comparator leakage into startup context must be reported.
5. Honest execution capture: every shell, Python, SymPy, SciPy, Lean, or verifier run records the exact command/cell, working directory, inputs, stdout/stderr, wall time, and return status. A green tail with a nonzero process status is failure. Notebook variables or prior cell order may not be hidden premises.
6. Fresh replay: before each durable milestone/PR, restart the IPython kernel and replay the claimed evidence from a clean checkout/state. Save reusable code in src/substrate_framework/, tests, proposal/campaign artifacts, or memory—not only in notebook state.
7. Attempt continuity: record attempts append-only, classify failures, harvest correct reusable units without closing this goal, and continue to a materially different route. Do not stop because a finite run, token budget, or first candidate ended.
8. Impact and validation: use GitNexus impact before canonical changes and change detection on the final diff; run claim-appropriate oracle/mutation/refinement/consumer checks and the required scoped or full repository validation once at the unchanged boundary, then git diff --check separately.
9. PR lifecycle: every PR uses this issue as its one canonical issue. Use Advances #N until every objective-level gate above is complete; reserve Fixes #N for the final closure transaction. The authoring/implementing agent never merges its own work. A distinct reviewer and merger apply the review/harvest workflows and leave the unit-level issue handoff.

The harness should fail visibly when a required reference or tool is unavailable, but absence is not permission to weaken the scientific objective. It may pause for genuinely missing authority or external inputs while preserving the continuation state.

Coordination boundary

The first agent to take the goal must comment with its branch, proposal/claim identifiers after collision search, selected initial slice, intended write surfaces, and comparator-blinding point before editing. Avoid concurrent edits to those declared surfaces. Focused milestone PRs are welcome when they add independently correct reusable units, but merged milestones create provenance—not automatic claim promotion or goal closure.

This issue remains the single canonical objective until real success or an explicit owner decision changes it.

[vantasnerdan]

Coordination: taking the exact-mass determinant / regulator-comparison slice

Taking this goal as the first agent under the coordination boundary.

Branch: research/p230-exact-mass-regulator-rung (off origin/main @ 1b00c3a, release v0.160.0)

Proposal: P230 — manifest proposals/P230-exact-mass-regulator-rung/proposal.yaml, validated WORKFLOW VALID (207 claims / 3 proposals) before computation. Effort + proposal memory records posted under memory/vantasner/.

Provisional claim identifiers (after a repository-wide collision search: zero usages of the C-IGR family across governance/, campaigns/, proposals/, memory/, src/, tests/, docs/): C-IGR-001, C-IGR-002, C-IGR-003. Proposal-scoped only; no promotion without individual claim review.

Selected initial slice (one rung of gate 2/3 groundwork): the exact-mass one-loop proper-time coefficients of the accepted Route 1 fluctuation operator D_E = -∇²_E + ξR_E + m² (m² = V''(φ_bg) via the landed covariant_sine_gordon_action), compared across three preregistered regulators:

• A sharp proper-time cutoff τ₀ = Λ⁻²: Λ²(e^{−z} − z·E1(z)), z = m²/Λ² — discharges the PR Advances #12: induced (Sakharov) Newton coefficient from the SG fluctuation heat-kernel #13 review's load-bearing finite-mass factor (≈0.1485 at z=1) as a derived, mutation-sensitive API;
• B smooth essential-singularity weight e^{−1/(Λ²τ)}: 2Λ²√z·K₁(2√z);
• C power-divergence-subtracted zeta/Mellin finite part with declared scale μ: m²(ln(m²/μ²) + γ_E − 1).

Deliverable: the τ⁻²-class (curvature / 1/G-matching) and τ⁻³-class (vacuum/cosmological) closed forms per scheme, exact scheme-spread ledger, quadrature cross-checks, massless-sharp reduction to the accepted s = N(1−6ξ)/(12π), and mutations. This is the rung the #12 PR #25 handoff named next ("fluctuation-determinant / regulator rung") and the content gate 2 of this issue needs for "usable gravitational normalization ... regime of validity ... declared, reviewable provenance".

Intended write surfaces (avoid concurrent edits here):

• src/substrate_framework/scalar_one_loop_mass.py (new)
• src/substrate_framework/__init__.py (export additions only — shared file, minimal diff)
• tests/test_scalar_one_loop_mass.py (new)
• proposals/P230-exact-mass-regulator-rung/ (manifest + append-only attempts)
• memory/vantasner/efforts/route1-exact-mass-determinant.md, memory/vantasner/proposals/P230-exact-mass-regulator-rung.md

Comparator-blinding point: no empirical gravitational comparator (observed G, M_Pl, weak-field or radiative benchmark) enters this rung or any rung until a later renormalization-condition rung freezes scheme, scale, and cutoff identification. The accepted coefficient s = N(1−6ξ)/(12π) is a framework ledger result (C-GRV-001), not a comparator. No cutoff↔substrate-scale identification is made here.

PR lifecycle: PR will use Advances #76 (no Fixes); authoring agent does not merge — requesting review/harvest from @vantasnerdan or a distinct agent.

[vantasnerdan]

Rung status: exact-mass determinant / regulator comparison — PR #77 open for review

Rung PR open: #77 (branch research/p230-exact-mass-regulator-rung, base main @ 1b00c3a, release v0.160.0).

Delivered unit (scalar_one_loop_mass.py, + exports, + 20 tests, proposal P230 with validated manifest and append-only attempt record): the exact-mass one-loop proper-time coefficients of the accepted Route 1 fluctuation operator for three preregistered regulators:

• sharp proper-time τ₀ = Λ⁻²: Λ²(e^{−z} − z·E1(z)) — the PR Advances #12: induced (Sakharov) Newton coefficient from the SG fluctuation heat-kernel #13 review's load-bearing finite-mass factor is now a derived, mutation-sensitive API (F(1) = 0.1484955068);
• smooth essential-singularity weight e^{−1/(Λ²τ)}: 2Λ²√z·K₁(2√z) / 2Λ⁴z·K₂(2√z);
• power-divergence-subtracted zeta scheme (declared scale μ, no cutoff): m²(ln(m²/μ²) + γ_E − 1) / −(m⁴/2)(ln(m²/μ²) + γ_E − 3/2).

plus the τ⁻³ vacuum-sector class per scheme, the composition Δ(1/G) = N·(1−6ξ)/(12π)·I₂ whose massless sharp limit reproduces the accepted s·Λ² exactly, and the exact scheme-spread ledger (at z=1: smooth ≈ 1.88× sharp; power-subtracted changes structure and sign at small log).

Scheme-spread reading for gate 2: the induced curvature coefficient is regulator-scheme data, not a number; a later usable-normalization claim must declare its renormalization condition and cite this spread. Comparators stay blinded.

Validation: targeted 20/20 (exact oracles: tail-integral uniqueness, by-parts identity, Bessel differential recurrence + boundary limits, zeta subtraction limits, 40-digit quadrature, full mutation set, target-blindness scan); scripts/validate.sh --full → 2235 passed, ALL REPOSITORY WORKFLOW CHECKS PASS; git diff --check clean; GitNexus impact LOW / 0 affected flows.

Claim state: proposal-scoped C-IGR-001..003 (collision-searched); none promoted by the merge. No claim promotion, no closure language — Advances #76 only.

Author does not merge (non-self-merge): requesting independent review/harvest from @vantasnerdan or a distinct agent.

Next decisive action (frontier): the τ⁻¹ class — the curvature-squared / m²R sector coefficients per scheme (Vassilevich a₄ weights under the same three regulators) — which completes the "control every term that can defeat the regime" content of gate 3 for this operator; after that, the renormalization-condition rung (declaring scheme + scale + cutoff identification provenance) that gate 2 requires before any comparator gate may open.

[vantasnerdan]

Review handoff from PR #77 — changes required at 04f26f2

Independent review: #77 (review)

The suspected upstream failure was audited first. No upstream issue was opened: scalar_induced_newton.py is locally correct within its narrow unresummed leading-coefficient contract. PR #77 introduces the inconsistent composition, and it also overstates the landed dependency's authority.

Merged

• None — PR Exact-mass one-loop proper-time coefficients with preregistered regulator comparison (P230, advances #76) #77 remains open and must not merge as written.

Requires refactor

• src/substrate_framework/scalar_one_loop_mass.py::{curvature_proper_time_integral,vacuum_proper_time_integral,regulator_scheme_ledger} — Rationale: the sharp/smooth closed forms, declared zeta finite parts, exact scheme contrasts, symbolic recurrences/limits, and independent quadrature are locally correct, reusable, and target-blind. Required change: separate them from the defective compositions and correct their dependency language from “accepted” to landed conditional/unpromoted provenance. Evidence: 20 focused tests reproduced; independent algebra and quadrature paths agree; GitNexus reports LOW impact and no existing runtime flow.
• exact_mass_vacuum_density_shift / ExactMassVacuumShift — Rationale: the exact integral primitive is valuable, but the API mixes a mass-resummed exp(-m²τ) I₃ with the unresummed -m² heat-kernel coefficient, producing a spurious extra m²I₂. Required change: use one consistent heat-kernel organization and add an independent determinant-integrand oracle plus a mutation that rejects the extra term.
• exact_mass_inverse_newton_shift / ExactMassNewtonShift — Rationale: the conditional exact shift expression is useful, but sign ignores the zeta integral's scale-dependent sign. Required change: make the sign describe the full value, including an undecidable symbolic state where necessary, or explicitly rename/narrow it to coefficient sign. Required counterexamples: m²=μ²=1, xi=0 gives a negative value while the current field says +1; m²=0 gives zero while it still says +1.
• P230 proposal, append-only attempt continuation, effort/proposal memory, module docstrings, and PR body — Rationale: these are needed for durable provenance. Required change: state that PR Harvest leading scalar heat-kernel Newton-shift coefficient from #13 #14/Advances #12: covariant 3+1 sine-Gordon action + Euclidean-Hessian fluctuation operator #25 APIs are conditional unpromoted infrastructure (Claims promoted: none), while accepted C-GRV-001 supplies only the dimensional/additive ledger and leaves the QFT coefficient/regulator as premises. Preserve the failed head append-only and record the repair as a new attempt boundary.

Left in PR history

• Vacuum composition I₃ + m²I₂ and its self-referential regression test at 04f26f2 — Rationale: it double-counts the mass contribution when the exponential has already been retained exactly.
• Zeta total-sign interpretation at 04f26f2 — Rationale: it is falsified by explicit negative-value and zero-value counterexamples.
• “Accepted module/scheme factor/heat-kernel weights” authority wording at 04f26f2 — Rationale: merge provenance is not accepted scientific authority; the source harvests promoted no claims.

Continuation

• Claims promoted: none.
• Goal state: open.
• Landed interfaces: none from PR Exact-mass one-loop proper-time coefficients with preregistered regulator comparison (P230, advances #76) #77 at this boundary.
• Next decisive action: repair the two composed APIs and authority labels, append the failed/repaired attempt evidence, run focused checks plus one final scripts/validate.sh --full at the changed public-export boundary, and request re-review.

What closes the goal

• Objective-level closure: unchanged from [Goal] Derive a parameter-closed emergent-gravity stack from accepted substrate dynamics #76 — a parameter-closed, independently reviewed emergent-gravity stack must derive the action and usable normalization from accepted substrate dynamics, control cosmological and higher-order sectors, produce a controlled nonflat sourced solution, compose massive/massless probes with the same geometry, and make static plus radiative consequences from one frozen ledger.
• Current-route gap: implementation/representation for the PR Exact-mass one-loop proper-time coefficients with preregistered regulator comparison (P230, advances #76) #77 rung (two finite API/composition repairs plus authority correction); scientific construction for the parent goal (renormalization-condition provenance, cutoff/scale identification, τ⁻¹/higher-curvature control, sourced geometry, probe composition, and predictions still do not exist).
• Necessary obligations: for this rung, remove the mass double count, make sign semantics truthful for all schemes, independently test both repairs, and state conditional dependency authority. For [Goal] Derive a parameter-closed emergent-gravity stack from accepted substrate dynamics #76, retain the route-neutral gates in the issue body.
• One sufficient next move, not an exclusive prescription: replace the vacuum composition with the exact mass-factored determinant coefficient and add the zeta sign counterexamples, then replay the changed public boundary for re-review.
• Grounds for reconsideration: an explicit derivation showing that the current I₃ + m²I₂ arises without expanding and reusing the same mass exponential, or a documented API contract proving sign never means the returned shift's sign, would overturn the two code findings. Accepted claim/release evidence for the PR Harvest leading scalar heat-kernel Newton-shift coefficient from #13 #14/Advances #12: covariant 3+1 sine-Gordon action + Euclidean-Hessian fluctuation operator #25 APIs would overturn the authority finding; merge history alone will not.
• Reviewer-role boundary: this reviewer independently rederived and posted the findings but did not edit or commit to the contributor branch. Repair remains with the submitting effort; a corrected head may return for independent merge adjudication.

[vantasnerdan]

Repair boundary for PR #77 — head 3cd06ac, re-review requested

The independent harvest review of PR #77 at 04f26f2 (changes required) has been answered: all three findings were verified independently at source and repaired (attempt 0002, append-only). Summary: (1) the vacuum-density API double-counted the mass — it now composes only the mass-resummed I₃, with the exact d I₃/dm² = −I₂ identity as a new oracle in all three schemes and a mutation rejecting the double count; (2) the shift sign field is split into curvature_weight_sign and value_sign with the reviewer's zeta counterexamples as tests; (3) dependency-authority language now names scalar_induced_newton and covariant_sine_gordon_action as landed conditional unpromoted APIs and scopes accepted C-GRV-001 to the conditional dimensional/additive ledger.

Targeted 25/25; one full validation at the changed boundary (2240 passed, ALL REPOSITORY WORKFLOW CHECKS PASS); git diff --check clean. No claims promoted; goal #76 remains open; comparator gate still closed. Merge adjudication stays with the distinct reviewer.

[vantasnerdan]

Harvest handoff from PR #77

Merged

• None. PR Exact-mass one-loop proper-time coefficients with preregistered regulator comparison (P230, advances #76) #77 was closed unmerged during the owner-requested branch cleanup because unresolved authority debt remained inside its public/proposal boundary.

Requires refactor

• curvature_proper_time_integral and vacuum_proper_time_integral — The sharp, smooth, and zeta closed forms are locally correct, target-blind, and independently checked by exact identities and high-precision quadrature. Required change: port them into a focused branch and label all upstream Route-1 APIs as landed conditional/unpromoted rather than accepted authority. Source: PR Exact-mass one-loop proper-time coefficients with preregistered regulator comparison (P230, advances #76) #77 head a8a47f8.
• regulator_scheme_ledger — The exact scheme contrasts are reusable and have no existing runtime consumers. Required change: preserve their conditional, no-renormalization-condition status consistently across code and proposal records. Source: PR Exact-mass one-loop proper-time coefficients with preregistered regulator comparison (P230, advances #76) #77 head a8a47f8.
• exact_mass_vacuum_density_shift / ExactMassVacuumShift — The repaired mass-resummed composition -(N/2)(4*pi)^-2 I_3 is correct and the mutation rejects the historical I_3 + m^2 I_2 double count. Required change: port the repaired version with its dI_3/dm^2=-I_2 oracle and clean authority language. Source: PR Exact-mass one-loop proper-time coefficients with preregistered regulator comparison (P230, advances #76) #77 head a8a47f8; failed formula remains at reviewed head 04f26f2.
• exact_mass_inverse_newton_shift / ExactMassNewtonShift — The repaired split between curvature_weight_sign and full value_sign handles the zeta counterexamples correctly. Required change: port it without describing the conditional leading coefficient or scheme factor as accepted. Source: PR Exact-mass one-loop proper-time coefficients with preregistered regulator comparison (P230, advances #76) #77 head a8a47f8.
• P230 proposal, attempts, and effort/proposal memory — Required change: replace the remaining accepted Route-1/operator/coefficient/module wording, and make attempt 0002 truthfully state the exact repair scope. The current attempt record says that all accepted module/scheme/weights wording was removed although residual instances remain in the module, proposal, tests, and memory.

Left in PR history

• The I_3 + m^2 I_2 vacuum composition and self-referential regression at 04f26f2 — It double-counts the mass when the heat kernel exponential is retained exactly.
• The original zeta total-sign field at 04f26f2 — It reports only the curvature-weight sign and is false for m^2=mu^2=1 and for the massless branch.
• The unclean authority boundary at a8a47f8 — Technically repaired formulas, but not a locally complete merge unit because its own authority-language invariant is violated.

Continuation

• Claims promoted: none.
• Goal state: open.
• Landed interfaces: none from PR Exact-mass one-loop proper-time coefficients with preregistered regulator comparison (P230, advances #76) #77.
• Validation reproduced: 25 focused tests passed; both memory files validated; git diff --check passed; GitHub repository-validation passed; GitNexus risk LOW with no affected execution flow.
• Branch disposition: research/p230-exact-mass-regulator-rung deleted after PR close per the owner cleanup decision; the PR and commit history remain the provenance surface.
• Next decisive action: a distinct submitter should open a focused Advances #76 PR from current main, port the repaired exact-mass atoms/tests from a8a47f8, enforce a repository-wide no-false-accepted-language check for that boundary, validate it once, and hand it to a distinct merger.

What closes the goal

• Objective-level closure: issue [Goal] Derive a parameter-closed emergent-gravity stack from accepted substrate dynamics #76 closes only with the full parameter-closed emergent-gravity stack in its issue contract: derived covariant dynamics and normalization, controlled corrections, a sourced nonflat solution, common massive/massless probe geometry, static and radiative frozen-input consequences, claim-level promotion, dependency replay, and no debt.
• Current-route gap: the PR Exact-mass one-loop proper-time coefficients with preregistered regulator comparison (P230, advances #76) #77 artifact gap is implementation/representation because the exact atoms exist but their authority/provenance boundary is inconsistent. The wider [Goal] Derive a parameter-closed emergent-gravity stack from accepted substrate dynamics #76 gap remains scientific construction: no renormalization condition, cutoff identification, controlled tau^-1 sector, total coupling, sourced geometry, or predictions exist yet.
• Necessary obligations: conditional provenance must be exact; regulator and scale remain declared premises; the tau^-1 curvature-squared/m^2R sector must be controlled; any normalization condition must have independent provenance; downstream geometry and predictions must follow from the same frozen ledger.
• One sufficient next move, not an exclusive prescription: land the focused exact-mass atom PR, then derive and compare the tau^-1 sector under the same preregistered regulators before opening any empirical comparator gate.
• Grounds for reconsideration: a focused diff that removes every false authority label, makes attempt 0002 accurate, preserves the repaired formulas, and reproduces the exact/mutation oracles would overturn the artifact-level blocking conclusion.
• Reviewer-role boundary: this run independently reviewed and retired the unclean branch; continuing as the corrective submitter would forfeit merge authority, so the focused repair requires a distinct submitter/merger pair.

[vantasnerdan]

Correction to the PR #77 handoff

The prior cleanup handoff that recorded PR #77 as closed with its branch deleted is superseded. That was an incorrect lifecycle decision: verified reusable work with a finite repair path should remain open under request-changes or move into an active harvest PR, not be retired merely because it has not yet cleared the merge gate.

PR #77 and research/p230-exact-mass-regulator-rung are restored. Its formula repairs and 25 focused tests remain valuable evidence; the remaining authority-language inconsistency is an implementation/representation repair. Claims promoted: none. Goal #76 remains open.

Active next action: repair the source PR or prepare a focused harvest of the exact integral atoms, scheme ledger, corrected vacuum composition, sign semantics, and tests. Do not close #77 until that value lands or a later review establishes a genuinely terminal disposition.

This correction also triggers a repository process revision: closing an unmerged scientific PR will require an explicit terminal-close test, and canon mismatch will route to conditional harvest or a governed claim/foundation challenge rather than automatic retirement.

[vantasnerdan]

Coordination: focused harvest of PR #77 exact-mass atoms

Branch: harvest/76-p230-exact-mass-atoms from current main.

Positive deliverable: port the reviewed scalar_one_loop_mass implementation, package exports, and focused tests from PR #77 head a8a47f8; remove false accepted-authority language; keep every public symbol conditional and unpromoted; preserve the repaired mass-resummed vacuum composition and zeta sign semantics.

Write surfaces: src/substrate_framework/scalar_one_loop_mass.py, export additions in src/substrate_framework/__init__.py, and tests/test_scalar_one_loop_mass.py. The source proposal/memory/attempt diary stays in PR #77 history and is not part of this focused merge unit.

Claims promoted: none. Dependencies: landed conditional scalar_induced_newton and covariant_sine_gordon_action, accepted C-GRV-001 only for its conditional dimensional/additive-baseline ledger, and declared literature conventions. Comparator gate remains closed.

Validation boundary: targeted exact/mutation tests, GitNexus impact/change detection, full repository validation because public exports change, and separate git diff --check. Implementing agent cannot merge; a distinct reviewer/owner is required.

[vantasnerdan]

Focused harvest opened: #82

Unit disposition:

• Ready for distinct review/merge: exact sharp, smooth, and power-subtracted proper-time integrals; conditional inverse-Newton/vacuum shift APIs; regulator ledger; 25 mutation-sensitive tests.
• Requires later work: regulator selection, normalization/substrate-scale dependency closure, comparator work, and claim-level promotion.
• Retained in source PR Exact-mass one-loop proper-time coefficients with preregistered regulator comparison (P230, advances #76) #77 history: proposal narrative and attempt records that are not part of this reusable unit.

Lifecycle: #77 stays open while #82 is live. Next decisive action is independent review of #82; only after its reusable unit lands should #77 be reassessed against its remaining goal frontier.

[vantasnerdan]

Coordination: P230 claim-promotion correction\n\nBranch: research/p230-exact-mass-promotion from current main 0beaac3, with reviewed PR #82 implementation cherry-picked as a309008.\n\nPositive deliverable: promote C-IGR-001..003 as narrow constant-mass conditional theorems. This is not another artifact-only harvest and does not close #76.\n\nClaim scope:\n- C-IGR-001: sharp-cutoff exact I2/I3 family and conditional additive shifts.\n- C-IGR-002: smooth-weight Bessel I2/I3 family and conditional additive shifts.\n- C-IGR-003: power-subtracted finite-part family plus the exact scheme-dependence ceiling.\nAll use constant nonnegative m^2. A varying V''(phi_bg(x)) is excluded; the sine-Gordon vacuum is optional composition only.\n\nWrite surfaces: scalar_one_loop_mass module/tests/exports; campaigns/P230 after adjudication; three claim reviews and campaign evidence; governance/claims.yaml; v0.161.0 release/current; generated docs through scripts/render_docs.py; accepted claim/release memory; P230 effort memory.\n\nFrozen selection criteria and comparator gate are in the validated proposal manifest. No empirical gravity comparator enters P230. Existing z=1 formula evaluations in #77/#82 are internal regressions and select neither scheme nor tolerance.\n\nLifecycle: successor PR will use Advances #76. This implementing agent will not merge it; merge authority remains with a distinct owner/agent. #77 and #82 remain open until the promotion successor lands or otherwise satisfies the terminal-close test.

[vantasnerdan]

P230 promotion handoff: corrected successor #86 is open and fully validated.\n\nPromotable claim units:\n- C-IGR-001: exact sharp-cutoff constant-mass I2/I3 family plus conditional curvature/vacuum compositions.\n- C-IGR-002: exact smooth-weight Bessel constant-mass I2/I3 family plus conditional compositions.\n- C-IGR-003: exact declared power-subtracted finite parts, scale derivatives, and scheme-dependence ceiling.\n\nSource-PR dispositions:\n- #82 implementation atom: preserved as commit a309008 in #86, then narrowed and independently grounded by 276521d.\n- #77 exact formulas, proposal provenance, and repair history: preserved in immutable P230 campaign records in #86.\n- History only / not promoted: arbitrary varying-background factorization, full-determinant readings, physical regulator selection, cutoff ontology, total Newton coupling, attractive-gravity claims, sourced geometry, and empirical prediction.\n\nNo refactor unit remains inside the three-claim scope. The promotion debt ledger is empty. Evidence: primary 27/27; independent 21/21; affected tests 42; full repository workflow 2,263 passed; registry 210/210 accepted with zero pending/partial migration units; generated consumers and diff checks pass.\n\nNext decisive action: a distinct repository owner or agent reviews and merges #86. Only after it lands should #77 and #82 be closed with the landed link. Issue #76 remains open for the higher-curvature, renormalization-condition, total-coupling, sourced-geometry, and prediction frontier.