refactor: derive recognition from typed evidence

[akcorca]

Closes #189

Summary

• replace raw domain and law prose scans with clause-scoped typed evidence
• hard-cut pack activation from arbitrary patterns to reviewed phrases and closed structural kinds in schema 7
• preserve explicit formula operation constraints without law-specific string branches
• rebuild the release WASM on the designated x86_64 Linux builder
• remove legacy domain/law paths detected by the architecture audit

Verification

• cargo test --workspace
• bun run pack:authoring
• bun run pack:conformance
• bun run legacy:audit (only v0.19 3/6: Index law dispatch for hundreds of packs #190 and v0.19 4/6: Centralize meaning decisions and resolution plans #191 owners remain)
• default bun run check passes through package/conformance/parity; the local RSS gate was noisy by <1 MiB and will be judged by CI

The strict corpus now exposes decisions formerly supplied by the removed contradiction-window heuristic. Those are intentionally not restored here; #191 owns their typed decision policy and the final epic gate.

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 3585446b-39f3-45cf-af52-e83062aac079

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}