Implement capability-oriented chat runtime

[anth-volk]

Fixes #260

Summary

• replace the previous linear orchestration with typed, reusable capabilities and tools
• preserve one authoritative typed conversation context across turns
• validate, resolve, and atomically apply model-proposed context changes
• add household, policy-reform, policy-information, society-analysis, chart, and follow-up capabilities
• add SQL-backed capability, context, artifact, and trace persistence with Alembic migrations
• expose structured private and public invocation activity through the optional debug setting
• remove superseded orchestration, event, prompt, and evaluation pathways
• prevent direct monetary facts from being silently discarded and use only accepted context facts during collective-income resolution
• add multi-turn regressions for every reported household conversation failure

Verification

• 445 backend tests passed; 15 skipped
• 9 multi-turn regression cases pass, including the conditionally enabled live Anthropic case
• backend type checking passed for 35 source files
• strict OpenSpec validation passed
• local backend health check and frontend HTTP check passed

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
policyengine-uk-chat	Ready	Preview	Aug 30, 2026 10:03am

[codecov]

Codecov Report

❌ Patch coverage is 80.77209% with 1036 lines in your changes missing coverage. Please review.
✅ Project coverage is 74.50%. Comparing base (536bd06) to head (29eb2c3).
⚠️ Report is 5 commits behind head on main.

Files with missing lines	Patch %	Lines
backend/capabilities/household.py	79.85%	90 Missing and 53 partials ⚠️
backend/conversation_context/reducer.py	64.21%	63 Missing and 44 partials ⚠️
backend/conversation_context/change_pipeline.py	73.65%	70 Missing and 33 partials ⚠️
backend/chat/capability_service.py	82.66%	45 Missing and 28 partials ⚠️
...ackend/conversation_context/variable_resolution.py	82.46%	35 Missing and 22 partials ⚠️
backend/capabilities/policy_reform.py	76.75%	33 Missing and 20 partials ⚠️
backend/capabilities/household_input.py	80.00%	34 Missing and 18 partials ⚠️
backend/conversation_context/models.py	85.67%	26 Missing and 26 partials ⚠️
backend/chat/context_coordination.py	66.16%	29 Missing and 16 partials ⚠️
backend/capabilities/society.py	77.63%	22 Missing and 14 partials ⚠️
... and 27 more

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #261      +/-   ##
==========================================
+ Coverage   68.00%   74.50%   +6.49%     
==========================================
  Files         109      153      +44     
  Lines        6827    11468    +4641     
  Branches     1390     1973     +583     
==========================================
+ Hits         4643     8544    +3901     
- Misses       1771     2189     +418     
- Partials      413      735     +322     

Flag	Coverage Δ
backend	81.76% <80.77%> (+0.46%)	⬆️
frontend	37.46% <ø> (+4.23%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
backend/capabilities/__init__.py	100.00% <100.00%> (ø)
backend/capabilities/artifacts.py	100.00% <100.00%> (ø)
backend/capabilities/compatibility.py	100.00% <100.00%> (ø)
backend/capabilities/input_resolution.py	100.00% <100.00%> (ø)
backend/chat/__init__.py	100.00% <100.00%> (ø)
backend/chat/activity.py	100.00% <100.00%> (ø)
backend/chat/events.py	100.00% <100.00%> (ø)
backend/chat/routes.py	77.27% <ø> (ø)
backend/chat/schemas.py	100.00% <100.00%> (ø)
backend/chat/turn_input.py	88.88% <100.00%> (+0.65%)	⬆️
... and 78 more

... and 5 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[github-actions]

Beta preview has been cleaned up because this PR was closed.