Ralph/prd 141 widget vertical refactor#388
Merged
Merged
Conversation
- US-011 (byte-equality snapshot tests) moved ahead of US-005 so the equivalence net guards every lift (US-005-010) in real time instead of landing after the US-010 hard-cut delete. - US-012 grep gate: drop PCRE (?!...) lookahead (incompatible with grep -E; macOS has no grep -P) for a portable two-pass POSIX ERE check. - US-009: Ralph validates the migration offline only (alembic --sql + py_compile + reviewer agent); staging apply + rowcount deferred to human.
Adds byte-equality tests against the US-004 fixtures that run through the US-003 shim end-to-end. Conftest AST-extracts the four proactive helpers from real chat.py and injects them as a fake api.widgets.chat module plus a fixture-bound GraphifyService stub, so the tests exercise production source without paying the full FastAPI/RAG import cost. The synthetic INBUILD-flavoured graph exercises every branch (FBT walk, collection/vendor fallback, multi-seed cart aggregation, by_vendor- overrides-FBT NetworkX undirected quirk). Now-green snapshot net gates the US-005/006/007/008/010 lifts — must stay green through every move. Validation: pytest orchestrator/integrations/ → 32 passed; py_compile clean on the two changed Python files. `from main import app` is environment-blocked on a missing camelot native dep (unrelated to this PR; same blocker on baseline) so was not re-run as part of US-011. Story: scripts/ralph/prd.json US-011 PRD: docs/PRDS/141-WIDGET-VERTICAL-AGNOSTIC-REFACTOR.md
…context_value into integrations/shopify/ First ACTUAL move out of orchestrator/api/widgets/chat.py. Creates integrations/shopify/context_fields.py hosting the field-mapping tuple and single-value formatter — verbatim copies, no value or behaviour changes. chat.py now imports both from the new module so _build_proactive_opener_message keeps rendering identical output through US-008. conftest.py and regenerate.py drop the two names from their AST-extract sets and seed them into the exec namespace via a direct import (context_fields has no heavy deps). US-011 byte-equality snapshot tests stay green (14/14). Story: scripts/ralph/prd.json US-005 PRD: docs/PRDS/141-WIDGET-VERTICAL-AGNOSTIC-REFACTOR.md
…egrations/shopify/ Move the single-seed FBT/collection/vendor walk verbatim from orchestrator/api/widgets/chat.py to orchestrator/integrations/shopify/widget_proactive.py. The shim's product-page path now calls the local function and only lazy-imports _build_proactive_opener_message from chat.py (until US-008). chat.py keeps the existing call site at the proactive_opener branch and imports the resolver from the new home; the function definition is gone from chat.py. Conftest drops the AST extraction for this name and test_widget_proactive's fake_chat fixture monkeypatches the local resolver so delegation tests still record calls. All 32 integrations/ tests stay green including the US-011 byte-equality snapshots. Story: scripts/ralph/prd.json US-006 PRD: docs/PRDS/141-WIDGET-VERTICAL-AGNOSTIC-REFACTOR.md
…rations/shopify/ Move the multi-seed cart-idle FBT walk verbatim from orchestrator/api/widgets/chat.py to orchestrator/integrations/shopify/widget_proactive.py. The shim's cart_idle path now calls the local resolver and only lazy-imports _build_cart_idle_opener_message from chat.py (until US-008). chat.py keeps the existing call site at the cart_idle branch and imports the resolver from the new home; the function definition is gone from chat.py. Conftest drops the AST extraction for this name and test_widget_proactive's fake_chat fixture monkeypatches the local resolver so delegation tests still record calls. All 32 integrations/ tests stay green including the US-011 byte-equality snapshots. Story: scripts/ralph/prd.json US-007 PRD: docs/PRDS/141-WIDGET-VERTICAL-AGNOSTIC-REFACTOR.md
…to integrations/shopify/ Move _build_proactive_opener_message and _build_cart_idle_opener_message verbatim from orchestrator/api/widgets/chat.py to orchestrator/integrations/shopify/widget_proactive.py. chat.py now imports both from the integrations module; widget_proactive's handle_widget_message calls the local builders directly (lazy api.widgets.chat imports dropped). Tests rewired: orchestrator/tests/test_widget_proactive_prd007.py imports the two builders from their new home; integrations/shopify/tests/test_widget_proactive.py's fake_chat fixture now monkey-patches the builders on widget_proactive (consistent with the resolver patches from US-006/007); conftest.py's AST extraction of chat.py is removed — only the GraphifyService stub remains. All 32 integrations tests pass including the byte-equality snapshot tests. Story: scripts/ralph/prd.json US-008 PRD: docs/PRDS/141-WIDGET-VERTICAL-AGNOSTIC-REFACTOR.md
…ify workspaces New Alembic migration sets settings.vertical='shopify' on every workspace whose JSONB settings already contain shopify_domain. Idempotent (WHERE NOT settings ? 'vertical'). Downgrade only strips vertical from rows we set it on. Lands in same PR as US-010 to keep INBUILD's PRD-007 + PRD-008-B dispatch routing intact. Validated OFFLINE only: py_compile OK, `alembic upgrade … --sql` renders upgrade + downgrade cleanly, `pytest integrations/` 32/32 green, migration-reviewer agent PASS. Staging apply + rowcount verification deferred to human before merge (per US-009 AC #7). Story: scripts/ralph/prd.json US-009 PRD: docs/PRDS/141-WIDGET-VERTICAL-AGNOSTIC-REFACTOR.md
Generic widget endpoint now resolves vertical from workspace.settings.vertical (default 'generic') and dispatches the message through PLUGIN_REGISTRY[vertical].handle_widget_message. The Shopify-specific helpers are no longer imported into chat.py — they stay private to integrations/shopify/. The product_context fallback in the open-callback-form SSE bridge dropped its productTitle/ productHandle reads; the LLM tool call is now the only source. Observability preserved: PROACTIVE_REWRITE and UNKNOWN_TRIGGER_REASON log lines rebuilt from plugin telemetry, now carrying vertical=. Plugin docstrings, context_note values, and test fixtures dropped "shim" wording — post-rewire the Shopify plugin owns the path end-to-end. US-009 backfill migration (cbc3b1a) ships in the same PR for the atomic deploy required by US-010 notes. Validation: pytest integrations/ → 32/32 green (snapshot byte-equality for PRD-007 + PRD-008-B openers held). grep -rE 'productHandle|productTitle|cartItems|shopify_' orchestrator/api/widgets/chat.py orchestrator/modules/context/sections/ returns zero matches. Manual smoke test on INBUILD is deferred to the US-020 canary (human-only). Story: scripts/ralph/prd.json US-010 PRD: docs/PRDS/141-WIDGET-VERTICAL-AGNOSTIC-REFACTOR.md
… surfaces Adds scripts/ci/check-no-shopify-in-generic.sh (POSIX ERE, two-pass: forbidden field keys + shopify_ prefix with plugin/integration/sync allowlist) and wires it into a minimal .github/workflows/ check-shopify-isolation.yml that runs on push to main/ralph and on PRs to main. Behaviourally verified offline: clean tree passes, injected productHandle fails Pass A, injected shopify_widget_id fails Pass B, and shopify_plugin/integration/sync pass through the allowlist. The 32 US-011 integration tests stay green. Story: scripts/ralph/prd.json US-012 PRD: docs/PRDS/141-WIDGET-VERTICAL-AGNOSTIC-REFACTOR.md
Adds the step-by-step guide a future engineer follows to ship vertical #2 without touching generic surfaces. Covers what a vertical is, when to add one, the copy-paste-modify recipe (with a ~30-line minimal plugin), the §12 MAY-NOT rules, snapshot + grep-gate test requirements, where admin UI goes, and skill registration. Points at integrations/shopify/widget_proactive.py as the working reference. Story: scripts/ralph/prd.json US-018 PRD: docs/PRDS/141-WIDGET-VERTICAL-AGNOSTIC-REFACTOR.md
…n; PRD-141 build complete (cross-repo + canary stories remain BLOCKED for human)
Adds docs/architecture/integration-walkthrough-barbershop.md — a worked
end-to-end walkthrough of adding a hypothetical "barbershop" vertical
(ClipShop UK booking site) against the Phase 1 plugin protocol. Mirrors
Shopify's package layout, handle_widget_message dispatch, and graph-
backed message builders to prove the abstraction is not Shopify-shaped
in disguise. Produces the PRD §13 target opener byte-for-byte ("Looking
at booking with Sarah for a fade — her next opening is Tuesday at 2pm.
Want me to hold it?"), routed via PLUGIN_REGISTRY exactly as Shopify is.
Catalogues seven Known Abstraction Gaps surfaced by the walkthrough:
no convention for partner-API calls inside the rewrite (Gap 1);
centralised PROACTIVE_TRIGGER_REASONS in chat.py (Gap 2); no missing-
skill detection at dispatch (Gap 3); duplicated _format_opener_context_
value (Gap 4); no page_context schema versioning (Gap 5); multi-location
overrides (Gap 6); sync Session in async protocol (Gap 7). Verdict:
Phase 1 protocol is vertical-neutral enough to ship; all gaps are
additive and defer to the PRD that adds vertical #2.
Story: scripts/ralph/prd.json US-019
PRD: docs/PRDS/141-WIDGET-VERTICAL-AGNOSTIC-REFACTOR.md
Ralph executable scope complete:
- US-001..US-012 done (Phase 0/1 scaffolding + lifts + CI gate)
- US-018, US-019 done (docs)
- US-013/014/015 — BLOCKED-CROSS-REPO (automatos-widget-sdk)
- US-016/017 — BLOCKED-CROSS-REPO (automatos-skills)
- US-020 — SKIPPED-HUMAN (operational canary)
RALPH_COMPLETE
Contributor
|
Warning Review limit reached
More reviews will be available in 56 minutes and 56 seconds. Learn how PR review limits work. Your organization has run out of usage credits. Purchase more in the billing tab. ⌛ How to resolve this issue?After more reviews become available, a review can be triggered using the We recommend that you space out your commits to avoid hitting the rate limit. 🚦 How do rate limits work?CodeRabbit enforces hourly rate limits for each developer per organization. Our paid plans include higher PR review limits than trial, open-source, and free plans. In all cases, reviews become available again over time. During sustained high-volume PR review activity, CodeRabbit may temporarily slow when the next review becomes available. Please see our Fair Usage Limits Policy for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (17)
✨ Finishing Touches🧪 Generate unit tests (beta)
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. Comment |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.