feat(btw): extract the rule-based classifier experiment

[YUZHEthefool]

Summary

Extract the prototype's deterministic rule classifier, its opt-in conversation hook, and tests as a standalone experimental candidate. Explicit /work admission bypasses classifier rules.

Related issue

Related: #133
Parent: #122
Design: #28

Stack

• Base: split/btw-skill-routes. Depends on PR #145.
• Sibling experiments on the same base: #147, #148. These are alternatives, not dependencies.
• Tested shared baseline: 54987c857178edae1c169f63f795fefbe169d3d0.

Behavior

Implement and evaluate only the rule-based candidate in its own experimental PR. Start with the prototype's frozen rules and run the shared dataset and capability fixtures before considering targeted changes.

Implementation notes

A separate model call, conversation-owned model routing, production default changes, and a multi-classifier switch framework are outside this candidate PR.

Experiment status

Base branch: split/btw-skill-routes. Depends on PR #145. This PR shows only this feature's changes relative to that dependency. The original prototype at 33ee103a62937db3e930c89ba47a648b75cc7772 is provenance; integration follows current master's owners and authorization contracts. The parent issue records review/integration order.

This is one of three sibling draft PRs on the same feature baseline. Candidate-specific missing work is listed below; the documentation-only experiment drafts do not claim an implemented classifier.

Validation

132 tests passed across nine relevant suites, including classifier rules, explicit admission, detached concurrent work, and real metadata conversion. The CI quality Pyright project passed using the Linux platform model, and the VitePress build passed. Final rebasing changed only the inherited six Chinese strings: all 9 metadata tests, Ruff and locale Prettier passed again; implementation/tests/docs remained identical to the validated candidate.

Commands run from the feature worktree (using the existing Python environment):

python -m pytest tests/unit/test_btw_task_classifier.py tests/unit/test_conversation_loop.py tests/unit/test_btw_work_loop.py tests/unit/test_btw_delivery.py tests/unit/test_config_metadata_i18n.py tests/unit/test_process_stage.py tests/unit/test_btw_status.py tests/unit/test_builtin_command_extensions.py tests/unit/test_core_import_smoke.py --basetemp=.tmp/pytest-r1-final-integration -q
pyright --project pyrightconfig.quality.json --venvpath D:/project/Zephirum --pythonplatform Linux
python -m pytest tests/unit/test_config_metadata_i18n.py --basetemp=.tmp/pytest-r1-final-metadata -q
# docs/
pnpm run docs:build

Python checks used the existing Python 3.14.1 environment with python -m pytest and task-local --basetemp directories. Local Node was 24.11.1; CI must validate the pinned toolchain. Relevant scoped formatting and Markdown checks were run. The root tracker records the final combined validation. No live model evaluation is implied by deterministic tests or documentation checks.

Compatibility and risk

Draft experiment. Rules remain disabled by default. Shared evaluation corpus, comparative measurements and acceptance thresholds are still required before product adoption; passing deterministic tests does not select this candidate.

Checklist

• A corresponding development Issue exists and this PR contains one feature or experiment.
• Tests and bilingual documentation accompany the change, or missing experimental implementation is explicitly described.
• Existing authorization, current APIs/configuration, Python 3.14+, and integrated documentation constraints are retained.
• No dependencies, credentials, generated site output, or fork publication claims are introduced.
• A human maintainer review and separate AI-assisted review have approved merging.

Repository references: AGENTS.md, GOVERNANCE.md, AI_POLICY.md, and the feat PR template.

Agent note

Codex prepared this PR at the user's request to split the original #28 implementation into reviewable features, with only classifier experiments marked as drafts. AI assisted extraction, current-path adaptation, tests, and English descriptions. The validation above was performed locally; remaining limitations are stated explicitly. This note is not a maintainer approval. No PR has been merged.

[xero-team-bot]

⚠️ This PR conflicts with its base branch and needs a rebase.

git fetch origin split/btw-skill-routes
git rebase origin/split/btw-skill-routes
# after resolving the conflicts
git push --force-with-lease

The needs-rebase label is removed automatically once the conflicts are gone.
(Xero-Team/AstrBot · detected by xero-bot)