feat(api,ui): showcase workspace-tagged scenario plans (#392)

[w7-mgfcode]

Summary

Implements showcase-workspace E3 (#392, umbrella #389): the two scenario plans the showcase pipeline saves gain workspace-aware tags, and the What-If Planner's saved-plans library gains a deep-linkable tag filter.

Backend (feat(api))

• DemoContext.workspace_name + pure _showcase_plan_tags() helper in app/features/demo/pipeline.py; both POST /scenarios save steps use it.
• Tag taxonomy (locked in the PRP): every pipeline-saved plan carries ["showcase", <kind>, "source:showcase"]; preservation="keep" runs additionally stamp workspace:<workspace_name|workspace_id> (label never empty; no workspace row → no workspace tag).
• The scenario_simulate_and_save step's data additively echoes the tags list sent.
• No migration, no schema change, no scenarios production code — scenario_plan.tags (JSONB, GIN) and GET /scenarios?tags= containment shipped in PRP-27.

Frontend (feat(ui))

• parseTagsParam() in frontend/src/lib/url-params.ts (trim/dedupe/cap-20, tested).
• Planner saved-plans library: clickable tag badges, removable active-filter chips + Clear, server-side filtering via useScenarios(tagFilter), distinct filtered-empty state, and ?tags= URL round-trip (/visualize/planner?tags=workspace:<name> deep-links one workspace's plans).

Tests

• Unit: _showcase_plan_tags (ephemeral / keep+named / keep+unnamed), updated step-body assertions, keep-run variants for both save steps; parseTagsParam vitest cases.
• Integration: workspace-tag containment round trip proving the umbrella criterion verbatim (2 tagged plans + control; ?tags=workspace:<label> returns exactly the 2; second tag narrows by AND containment).

Out of scope (per PRP)

• The agent-HITL plan stays untagged — SaveScenarioRequest carries no tags field; threading workspace context into the agent session is a cross-slice change deferred to a future epic.
• Tag editing/management UI.

Validation

• ✅ ruff check + ruff format --check + mypy --strict + pyright --strict
• ✅ Unit: 1966 backend passed; 336 frontend vitest passed; pnpm lint clean
• ✅ Integration: full -m integration suite green on a fresh DB (268 passed; seeder destructive tests need the pre-existing ALLOW_DESTRUCTIVE_TEST_DB opt-in locally, 16/16 green with it)
• ✅ Level-4 smoke: showcase_rich keep run → GET /scenarios?tags=workspace:<name> returns exactly the 2 plans with the full tag set
• ✅ Browser dogfood (Playwright, real Chromium): badge click narrows 3→2 rows server-side, chip + Clear render, URL carries ?tags=, deep link pre-applies the filter in a fresh tab, Clear restores the full list, filtered-empty state distinct from no-plans state

Closes #392

[sourcery-ai]

Sorry @w7-mgfcode, you have reached your weekly rate limit of 500000 diff characters.

Please try again later or upgrade to continue using Sourcery

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

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: defaults

Review profile: CHILL

Plan: Pro

Run ID: fc639284-ec16-4092-a69b-4b36e6958939

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

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat/showcase-workspace-tagged-plans

---

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 and usage tips.}