Repository navigation
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
… 58/257; final-tranche blocks 4–5 closed) (#421) Blocks 4 and 5 adopted (ledger v1.91, both audits CALIBRATED at 100%); Q5 seals document_071 (q2009), document_074 (fam2013), document_079 (q2019) and document_001 (fam1968); milestone entry covers the five open law rulings and the table-of-record materialization sweep. Facts from program-board lines 577–652 and the staged reconciled reports; seal dates taken from run ids because the board's time labels drifted after ~20:00. Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
…/257 = 77.8% of annotated; block 6 folded, back-catalogue reconciled) (#422) Block 6 adopted (ledger v1.92; fourth consecutive CALIBRATED audit, first with a correction); Q5 seals document_003, 017, 023, 027, 019 (doc_029 sealed then withdrawn on adversarial review, so the counter is 63, reconciled across three instruments); the Q5 unit now states both denominators (81 annotated questionnaires; 257 fixed source domain of which 43 can never seal). Milestone entry records the sentinel-catalog finding, the ratification question, the atomisation coverage finding (replicated), and the materialization closure. Facts from program-board lines 651–74x and the staged reports; seal dates from run ids. Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
…dger v1.97; Q5 63/257 = 77.8% of annotated) (#423) * Progress dashboard: 2026-09-05 afternoon refresh (purpose 100.00%, ledger v1.97 terminal fold; blocks 7–10 adopted; Q5 63/257 = 77.8% of annotated; decision package verified sound with named defects) The purpose census closes at R20815 = 20,815/20,815 after the block-10 audit returned CALIBRATED with an empty manifest (ninth consecutive). detail.json adopts blocks 7, 8 (folded in two parts after a scope error), 9 and 10 with their audit facts; progress.json carries the terminal ledger identity, the two-denominator Q5 framing, the corrected atomisation share (74.05%), the pinned reproducibility framing, and the decision-package verification verdict. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * Dashboard refresh: apply the 127-claim verification (7 wording refutations, block-10 audit source filename, 101 distinct sampled ranks) An Astra verification lane checked every figure and claim in the new milestone and block rows against the campaign artifacts: 118 verified, 7 refuted, 2 not established. All nine are wording precision (ordering criterion, five-item claim, cross-host reproducibility, who restated the extractor build, open questions, block-9 manifest items, block-10 sample count) plus one wrong source filename; no adopted number changes. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
…d-2 findings) (#424) * docs/plans: graph-port Plan A revision 5 (folds gate 20260905-154850 round 2: image identity and environment values in the TCB; latency per event class) Third gate 20260905-154850-plan-5cfd8532 (Fable main + Astra peer) on revision 4 requested two changes. This revision folds both: the hermetic-image alternative binds the complete image identity and the resolved semantic-environment values into the TCB with old-namespace rejection before any cache read (A2, A2a), and the H2 latency population is defined per event class with an explicit timestamp source, harvest-to-shadow-complete kept separately for harvest events (§4). No scope change; shadow-only and nonauthority as before. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs/plans: graph-port Plan A revision 6 (folds gate round 3: per-event-type applicability map and explicit absence semantics for the legacy comparison) Round 3 of gate 20260905-154850-plan-5cfd8532 found that the universal legacy comparison demanded a canonical report and T bytes for every event although voided attempts may never have produced either. Revision 6 adds a frozen applicability map per event class, an ABSENT-BY-DISPOSITION marker that satisfies A-2 and counts toward EVIDENCE-COMPLETE where the map permits it, a MISSING REQUIRED EVIDENCE failure where it does not, and fixtures for both cases. No scope change. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs/plans: graph-port Plan A revision 7 (folds gate round 4: applicability per report-derived and T-derived axis; report-bearing void fixture) Round 4 of gate 20260905-154850-plan-5cfd8532 found the revision-6 applicability map all-or-nothing: an attempt voided after its canonical report but before its T had no rule preserving the report comparison. Revision 7 partitions the legacy fields into report-derived and T-derived axes, declares applicability per axis from the authoritative disposition, keeps the report comparison for report-bearing voids, and adds the fixtures. No scope change. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs/plans: graph-port Plan A revision 8 (TCB pin to microcosm b55a583f; consume the core's store/codec/artifact facilities; string gate outcomes) Revision 7 was agreed by gate 20260905-154850-plan-5cfd8532. Revision 8 amends the A2 TCB pin from 6022e4f5 (which lacks register_bytes, raw-bytes-v1 and the typed ArtifactType/ArtifactInput/ArtifactOutput edges) to the commit the Microcosm US session froze for its interface check, b55a583f (memo SHA 485ea3d3…, receipt SHA 8d780c99…), recorded as a new TCB with its publication state; binds A-1 to the core's facilities instead of reimplementing them; and adds the string-outcome rule for gate kernels. Submitted to a fresh gate. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs/plans: graph-port Plan A revision 9 (folds gate 20260905-162709 round 1: all-string gate outputs, requires_decisions as a decision-name contract, per-fixture expected states) Three runtime-contract findings against the pinned executor at b55a583f: the exception path rejects any non-string owned output (executor.py:130); requires_decisions is matched against supplied decision names, not gate ancestry (executor.py:219-249); a missing required source is refused before any kernel executes (executor.py:1922-1924). Revision 9 requires all-string owned outputs on gate kernels with a mixed-output fixture, freezes and validates the decision tuple and the gate set independently with warm-replay checks, and scopes A-2's execute/hit assertions to each fixture's frozen expected state (executed, refused, unreached). Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs/plans: graph-port Plan A revision 10 (capacity amendment: upward-routed A-1/A-2 lane when Claude is capped, under ceremony precedence) Requested by the Microcosm US session under Max's steering to maximise parallel work. A5 and the matching §5 do-not now permit the isolated A-1/A-2 engineering lane to route upward to Astra on a Codex home when every Claude lane is capped by return code, subject to ceremony precedence on that home (no build while an Amendment-20 ceremony step runs there), a recorded window state and routing decision at dispatch, and the Claude path as the default whenever it frees first. No scientific, authority or scope change. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs/plans: graph-port Plan A revision 11 (folds gate 20260905-164552 round 1: blocking artifact edges, executed-failed vs unreached, both dependency-type fixtures) At the pinned executor a failed string verdict does not block downstream execution; only an unavailable typed artifact or an unreached ancestor does (_blocked_by). Revision 11 requires every gate that must stop execution to emit a typed artifact its dependents consume (transport gate -> reconstruct_T and seal gate; seal gate -> RELEASE), distinguishes EXECUTED-FAILED from UNREACHED in the frozen expected states, forbids a RELEASE that executes after a seal-gate failure, and requires cold/warm fixtures for both dependency types. The revision-10 capacity amendment drew no finding and is unchanged. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
…hed; referee questions; figure corrections; Plan A rev 11 merged) (#425) * Progress dashboard: 2026-09-05 evening refresh (C′ re-freeze landed and asserted; S′ dispatched; seven removed / twelve added sentences to the referees; 79-line and 126-id corrections; Plan A rev 11 merged, build running; kit r8b) * Evening refresh: add the Axis-C uniform pass (14.1% source-fidelity bound; 19.0% composite union) and the closure of both errata-named Q5 surfaces
…thheld ratification of C′; ceremony halted at the referee step; Plan A review, child build, revision 12 in gate) (#427)
…1fe7fe review; per-class staged rule; TCB digest and absent-digest fixes required) (#426) * docs/plans: graph-port Plan A revision 12 (two A-1 scopes after the 41fe7fe review; per-class staged rule; TCB digest and absent-digest fixes required) * Plan A revision 12: kernel totality contract (A2a) from the shadow child's measured counterexamples * Plan A revision 12: A-5 requires A-1 at live scope (gate round-1 finding folded)
…3 pass 1 dispatched) (#428) * Progress dashboard: 2026-09-06 morning refresh (A-1 accepted at LIVE scope on run 4's head; A-3 pass 1 dispatched; ceremony still halted at the referee step) * Progress dashboard: fold the A-3 pass 1 landing (accepted; A4 custody gap surfaced; H2 not yet reachable) * Progress dashboard: correct the A-3 custody finding (A4 is per-source; addendum rows, not a plan amendment)
…he record states it; measured forecast to DynaSim-class capability; levers in motion; A-3 pass 2; Q5 L1188 retraction; outage attribution corrected) (#429)
…rst day back: eight shadow branches; gate-3 lock discipline; Q5 and ceremony decisions re-priced by measurement; immigration fork) (#430)
…easured and fixed; gate-3 fourth sitting, second verification NOT VERIFIED on typed sentences, floors v4 and the chained third verification; A-3 pass 4 accepted; lane capacity as the binding constraint) (#431) Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
…READY by both referees, ruling now Max's; mortality gate floors v3 → referees → v4 → verification VERIFIED → threshold binding; the seat's overnight stop and its fixes; machine sleep; the lane drought; A-3 pass 4) (#432) Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
…2 → verification VERIFIED → threshold binding landed; Fable cap and the referee-loop defect fixed; the seat's third overnight stop and the Opus interim seat; kit r9 gate failed with repairs, r10 on Astra; orchestrator's midnight relaunch) (#433)
…closed to a ruling package — three packages with Max; kit r10 installed on concurrence with the re-render requirement; A-3 pass 5 censored on Astra, pass 5b accepted on a Claude host; the battery afternoon; the orchestrator's handover and first deliberate rotation; the token-hygiene finding and patch) (#434)
…in force and the orchestrator on Fable; A-3 pass 6 prepared early, granted on a measured lane, dry-run first (D15: sandbox mode is host capability), run on Fable at farness and accepted: 41 new events, 113 cumulative, zero discrepancies; the lane's bound misstatement self-corrected against the commit clock; the farness five-hour limit pausing the orchestrator until eleven) (#435)
This was referenced Sep 9, 2026
Draft
Contributor
Author
This branch was successfully deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The optional population graph reuses the existing M6 mortality fitter and mortality/ageing steps through typed model artifacts. It fits a historical exposure panel, applies the model to a separate starting population with stable person-ID draws, and retains original and next-period observations with explicit lineage and mass accounting. Separate source declarations keep held-out outcomes out of fitting and application.
Malformed held-out input now produces a retained execution manifest and a diagnostic report naming the failed evaluation gate and its original exception. This works on cold and cached failures; the CLI exits nonzero and marks unevaluated metrics as
not_evaluated. Regression tests also cover direct-model cutoff parity and nondefault experiment, replicate and seed coordinates.Depends on Microcosm #873. The graph extra still pins graph and Frame to
3ff92b0aea14407d09479bff5623dc7d1a92d008and requires Python 3.13+. Default Dynamics imports support Python 3.10–3.14. The existing CI job requires typed capabilities and runs the scoped integration on Python 3.13 and 3.14. Core interface review remains required before release. The source consolidation in Microcosm #893 overlaps #873 and #885; reconcile those contracts before landing either integration. This Dynamics PR retains its tested exact core pin pending separately reviewed compatibility work.Validation on September 9, from current master
507f71eplus this integration:git diff --checkpassed. The tier documentation now matches the manifest's 1,596 unit / 5,791 total entries; full-suite collection was not rerun locally because normal collection can import staged-source-dependent modules.These checks establish synthetic engineering behavior only. Household accounting, births, immigration, repeated-year projection and scientific mortality validation remain outside this slice. The existing M6 loop, Q5 campaign, candidate registries, locked scientific gates and committed evidence artifacts are unchanged. The previously reviewed birth-evidence exclusion remains limited to the six exact graph paths. This draft updates the existing PR with the previously unpushed diagnostics commit and current master; it does not certify or publish a population release.
Annual projection is developed separately in stacked draft #436, which preserves this PR's exact graph/Frame pin and adds repeated-year synthetic execution.