feat(forecast): drive selection from causal memory

[koala73]

Summary

• build a canonical causal graph over reportable effects and interactions, with continuity against prior runs
• feed prior memory and causal edges back into publish selection, and reserve early slots for persistent memory-backed situations
• add tolerant combined/scenario JSON extraction plus parse diagnostics, and clean up replay/memory watchlist wording

Validation

• node --check scripts/seed-forecasts.mjs
• /Users/eliehabib/Documents/GitHub/worldmonitor/node_modules/.bin/tsx --test tests/forecast-detectors.test.mjs tests/forecast-trace-export.test.mjs
• npm_config_cache=/tmp/worldmonitor-npm-cache npm exec --yes @biomejs/biome@2.4.7 -- lint scripts/seed-forecasts.mjs tests/forecast-detectors.test.mjs tests/forecast-trace-export.test.mjs

[vercel]

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

1 Skipped Deployment

Project	Deployment	Actions	Updated (UTC)
worldmonitor	Ignored		Mar 20, 2026 9:11pm

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.