Offer 2036 as a US year in economy options and the v2 catalog

[DTrim99]

Fixes #3927

Summary

• This is the policyengine-api part of Allow US calculations in 2036 #3926: it lets US users pick 2036.
• policyengine_api/country.py: adds dict(name=2036, label="2036") to the US time_period options. The app's year picker, date ranges and budget-window starts all read this list. The UK is unchanged (2024–2030).
• policyengine_api/data/v2/catalog/extraction.py: METADATA_TIME_PERIODS["us"] becomes range(2036, 2021, -1).
• docs/canonical-spm.md: the artifact horizon becomes 2022–2036.

Draft until the bundle with a 2036 SPM forecast is ready.

• The canonical SPM forecast in the current bundle covers 2022–2035. A 2036 request fails with SPM_YEAR_UNAVAILABLE.
• Merge this in the same release as the PolicyEngine bundle update that carries spm-calculator 1.1.0 and its 2022–2036 forecast. Merging earlier would show 2036 in the app while every 2036 run fails.
• The v2 catalog stores the year list per published model version, and publication reconciliation compares staged and stored lists. So the new list should go out with a new bundle version.

Budget windows already allow end years through 2099 (BUDGET_WINDOW_MAX_END_YEAR). The worker reads its year boundary from the installed SPM provider. Neither needs a change here.

The SPM tests that use 2036 as an unavailable year (test_spm_settings.py, test_spm_year_worker_polling.py, test_budget_window_cache.py) run on mocked forecasts with years=(2022, 2035). They don't depend on the installed artifact, so they're unchanged.

Testing

• ruff format and ruff check are clean on the changed files.
• No unit test pins the US year list. I didn't run the suite locally; CI covers it.
• After the bundle bump, on staging:
  • GET /us/metadata lists 2036.
  • A national 2036 economy run completes, with "2036" in its SPM receipt.
  • budget_window:2027:10 completes.
  • 2037 still returns SPM_YEAR_UNAVAILABLE.

🤖 Generated with Claude Code

https://claude.ai/code/session_01LqCGPgxDukji8Bhs57c5rH

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 87.58%. Comparing base (5c448d7) to head (b242eb4).

Additional details and impacted files

@@             Coverage Diff             @@
##           master    #3928       +/-   ##
===========================================
+ Coverage   45.25%   87.58%   +42.33%     
===========================================
  Files         183      198       +15     
  Lines       11085    12223     +1138     
  Branches     1907     2161      +254     
===========================================
+ Hits         5016    10706     +5690     
+ Misses       5647      921     -4726     
- Partials      422      596      +174     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.