Add ingredient metadata to simulation API calls

[anth-volk]

Summary

Fixes #3111

Adds ingredient metadata to society-wide economic calculation calls so that Logfire can trace errors back to specific policy IDs.

Changes

Modified policyengine_api/services/economy_service.py in the _handle_create_impact method to include a _metadata field in the simulation API call parameters:

sim_params = sim_config.model_dump()
sim_params["_metadata"] = {
    "reform_policy_id": setup_options.reform_policy_id,
    "baseline_policy_id": setup_options.baseline_policy_id,
    "process_id": setup_options.process_id,
}
sim_api_execution = simulation_api.run(sim_params)

How it works

1. The Logfire span in the simulation API (app.py) captures input_params=params before validation
2. The _metadata field is included in that capture
3. SimulationOptions.model_validate() silently ignores the extra _metadata field (Pydantic v2 default behavior)
4. No changes required to the simulation API itself

Benefits

• Enables direct correlation between Logfire errors and database records
• Helps debug issues like the Virginia policy bug where parameter values were incorrectly stored
• Minimal code change with high debugging value

Test plan

• Verify simulation API calls still work correctly
• Check Logfire to confirm _metadata appears in the run_simulation span

🤖 Generated with Claude Code

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 70.56%. Comparing base (647768b) to head (250f65e).
⚠️ Report is 2 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #3112      +/-   ##
==========================================
+ Coverage   70.53%   70.56%   +0.02%     
==========================================
  Files          55       55              
  Lines        2386     2388       +2     
  Branches      339      339              
==========================================
+ Hits         1683     1685       +2     
  Misses        642      642              
  Partials       61       61              

☔ View full report in Codecov by Sentry.
📢 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.