Repository navigation
Add certified-bundle replay pipeline for recent OBR fiscal events - #157
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
…the lane stopped (#156) The track 1 lane hit its wall-clock limit with these files in the worktree and not yet pushed: completed Autumn Budget 2024 artifacts (private school VAT, the SDLT additional dwelling surcharge, the Winter Fuel means test) and the files updated alongside them. Committed as found; not yet reviewed. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…pted Modal run (#156) 33 source rows now have computed counterparts across the five partial measures. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…dle (#156) Autumn Statement 2023, Spring Budget 2024, Spring Statement 2025 and Spring Budget 2023 ran on Modal, passed the adoption validation and are staged and compared. All five recent events now have numerical comparisons: 130 computed source rows across 13 partial measures. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
Independent review (GPT-6.1 Sol, read-only): NEEDS-FIXES. Full text: Blocking
Other findings
Divergence causes found
A fix lane is running: |
Review (GPT-6.1 Sol): B1 the UC health element construction was inert, now refreshed after the reform; B2 reform dates and the fiscal-year conversion, with docs corrected and the Annual Allowance computed from 2023-24; B3 inert constructions now block numerical completion; M1 UC counterfactual on the 3.8% CPI, naming policyengine-uk#2239; M2 the gap split shown everywhere; M3 the divergence diagnoses with policyengine-uk#2237 and #2238; L1 notes. Adds LHA reset, UC childcare caps and Class 2 abolition constructions. Reran Spring Statement 2025, Spring Budget 2023 and Autumn Statement 2023 on Modal: 165 computed rows (was 130). 176 replay and 324 extended tests pass; accounting identities hold. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
CI's ruff format check flagged four files. The re-review (MERGE-SAFE at 133613e) suggested 'Registered construction replay complete' over 'Numerical replay complete', since the replays cover registered partial constructions only; summaries regenerated. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
Merge audit: GPT-6.1 Sol review NEEDS-FIXES at 2c231b4 (B1-B3, M1-M3, L1); fixes in 133613e; re-review MERGE-SAFE at 133613e; 35549bf adds only ruff format and the reviewer's suggested status label. Checks green at 35549bf; MERGEABLE; no app or DB-build changes, so nothing is published. Squash-merged under Max's merge-on-gates rule. |
Refs #156 (track 1). Replays recent OBR-costed fiscal events on the certified bundle (populace-uk-2023, policyengine-uk 2.89.2), generalising the Autumn Budget 2025 port (#136).
What's here
build_uk_event_registry,compute_uk_event,stage_uk_event,compare_uk_event, the Modal UK adaptermodal_uk_eventandadopt_uk_event.docs/uk_replay/RECIPE.mdandYEARS.md.construction_pendingand 76 model or data gaps.Review
ruff formatand the reviewer's suggested label ("Registered construction replay complete").Divergence causes (from the review)
Invariants
-(reversal − certified).Tests: 176 replay and 324 extended tests (fix lane); 128 comparison, compute and registry tests after the label change.
The PR does not touch the app or the DB build, so merging publishes nothing.
🤖 Generated with Claude Code