mirror of
https://github.com/CopilotKit/CopilotKit.git
synced 2026-09-14 16:26:20 +08:00
Tyler Slaton
ef1ca9808c
Result of 10 parallel QA agents auditing all 30 active demos against
langgraph-python (north-star). Each agent ported drift back to LP-verbatim
across three axes:
1. Agent layer
- tool_rendering_common.py: rebuilt to LP's surface — get_weather,
search_flights(origin, destination), get_stock_price, roll_d20,
roll_dice. Removed the ADK-only query_data.
- tool_rendering_*_agent.py (4 variants): ported LP's travel/concierge
prompt; reasoning-chain variant got LP's chain-two-tools prompt.
- beautiful_chat_agent.py: ported LP's per-tool system prompt; added
manage_sales_todos / get_sales_todos / generate_a2ui; dropped the
redundant schedule_meeting (frontend HITL handles it).
- open_gen_ui_agents.py: ported LP's full SYSTEM_PROMPT for both
variants, including the Websandbox.connection.remote.* contract
for the advanced sandbox demo (was `window.sandbox.*`, which the
LP frontend's Websandbox bridge silently no-ops).
- byoc_agents.py: fused LP's hashbrown + json-render prompts so the
single ADK byoc_agent emits both wire shapes. Aliases exported for
a future per-route split.
- declarative_gen_ui_agent.py: ported LP's a2ui_dynamic SYSTEM_PROMPT.
- a2ui_fixed_agent.py: picked up LP's #4734 regression guard
("exactly ONCE", "do NOT call again").
- agent_config_agent.py: rewrote to read useAgentContext (was
state["config"]); reconciled schema to LP's 3-field camelCase
{tone, expertise, responseLength} with LP's value enums.
- subagents_agent.py: dropped the "running" placeholder; returns
plain str so the LP-verbatim frontend's `result?.trim()` works.
- hitl_in_app_agent.py / hitl_in_chat_book_call_agent.py: prompts +
tool-result shape ({approved, reason}) aligned to LP.
- AGUIToolset() added wherever it was missing on the bespoke agents
(multimodal, mcp_apps, a2ui_fixed) so frontend-registered tools
reach the model.
2. Dedicated runtime routes
- copilotkit-multimodal/route.ts (new) — mirrors LP shape with
ADK's HttpAgent + AGENT_URL pattern.
- copilotkit-agent-config/route.ts (new) — same pattern.
- copilotkit-mcp-apps/route.ts — refreshed.
3. Frontend ports (ADK frontend brought to LP-verbatim where it had
drifted from the parity blitz state)
- tool-rendering family (4 demos): full re-port — WeatherCard,
FlightListCard, StockCard, D20Card, ReasoningBlock, CatchallRenderer,
suggestions, and the page wiring with all useRenderTool /
useDefaultRenderTool / reasoningMessage registrations.
- a2ui-fixed-schema, mcp-apps, multimodal: full frontend re-ports
with their _components/ Tailwind primitives.
- frontend-tools, frontend-tools-async, agent-config: ported LP's
component structure (separate Background, NotesCard with query_notes,
config-context-relay).
- shared-state-read, shared-state-read-write, readonly-state-agent-context:
ported LP's demo-layout + _components + suggestions. recipe-card.tsx
pulled directly from LP (one QA agent had adapted to Unicode glyphs
thinking ADK lacked lucide-react — it doesn't, after the parity blitz).
- shared-state-streaming, subagents, hitl-in-app: ported LP's
DocumentView / supervisor-activity / TicketsPanel structure.
hitl-in-app/page.tsx pulled directly from LP to keep the hyphenated
agent slug aligned with the renamed registry key.
- auth, hitl-in-chat: ported LP's SignInCard-first auth UX and the
time-picker Tailwind port.
- prebuilt-popup: pulled LP's main-content + suggestions split.
4. Test fixtures
- 30 tests/e2e/<slug>.spec.ts ported from LP, several overwriting
stale stubs (shared-state-streaming, subagents, auth, hitl-in-chat,
shared-state-read, agent-config).
- 30 qa/<slug>.md ported from LP with ADK env-var and registry
references substituted (GOOGLE_API_KEY, AGENT_URL, registry.py).
- QA3's byoc-hashbrown / byoc-json-render specs renamed to
declarative-hashbrown / declarative-json-render with internal
URL references substituted (the orchestrator pass had already
renamed the demo dirs + manifest entries).
Frontend changes from QA agents were filtered: kept where they ported
LP-verbatim into ADK, replaced with direct LP pulls where the agent
had made ADK-specific adaptations (one Unicode-glyph case, one
stale-registry-slug case).
Not touched per blitz rules: shared_chat.py, registry.py, manifest.yaml,
src/app/api/copilotkit/route.ts.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
62 lines
4.6 KiB
Markdown
62 lines
4.6 KiB
Markdown
# QA: Shared State (Read + Write) — Google ADK
|
|
|
|
## Prerequisites
|
|
|
|
- Demo is deployed and accessible at `/demos/shared-state-read-write` on the dashboard host
|
|
- Agent backend is healthy (`/api/health`); `GOOGLE_API_KEY` is set on Railway; agent server is reachable at `${AGENT_URL}/shared-state-read-write`
|
|
|
|
## Test Steps
|
|
|
|
### 1. Basic Functionality
|
|
|
|
- [ ] Navigate to `/demos/shared-state-read-write`; verify the page renders within 3s with the preferences + notes cards in the main column and the `CopilotSidebar` open by default on the right (cards stack vertically once the viewport drops below `xl` / 1280px)
|
|
- [ ] Verify `data-testid="preferences-card"` is visible with heading "Your preferences"
|
|
- [ ] Verify `data-testid="notes-card"` is visible with heading "Agent Scratch pad" and empty-state `data-testid="notes-empty"` reading "the agent will make observations about you and note them here!"
|
|
- [ ] Verify the chat input placeholder is "Chat with the agent..."
|
|
- [ ] Verify all 3 suggestion pills are visible with verbatim titles: "Greet me", "Remember something", "Plan a weekend"
|
|
- [ ] Send "Hello" and verify an assistant text response appears within 10s
|
|
|
|
### 2. Feature-Specific Checks
|
|
|
|
#### UI Writes -> Agent Reads (preferences via `agent.setState`)
|
|
|
|
- [ ] Type "Atai" into `data-testid="pref-name"`; verify `data-testid="pref-state-json"` updates to include `"name": "Atai"`
|
|
- [ ] Change `data-testid="pref-tone"` to `formal`; verify the JSON preview reflects `"tone": "formal"`
|
|
- [ ] Change `data-testid="pref-language"` to `Spanish`; verify the JSON preview reflects `"language": "Spanish"`
|
|
- [ ] Click the `Cooking` and `Travel` interest pills; verify both show the selected style (border `#BEC2FF`, bg `#BEC2FF1A`) and the JSON preview's `interests` array contains both entries
|
|
- [ ] Send "What do you know about me?"; verify within 10s the assistant reply references the name "Atai", a formal tone, Spanish, and the Cooking/Travel interests (the `_inject_preferences` before-model callback injects these into the system prompt each turn)
|
|
- [ ] Click the "Plan a weekend" suggestion; verify the reply is tailored to the selected interests
|
|
|
|
#### Agent Writes -> UI Reads (notes via `set_notes` tool)
|
|
|
|
- [ ] Click the "Remember something" suggestion (sends "Remember that I prefer morning meetings and that I don't eat dairy.")
|
|
- [ ] Within 15s verify `data-testid="notes-list"` appears in the notes card and contains at least 2 `data-testid="note-item"` entries mentioning "morning meetings" and "dairy"
|
|
- [ ] Verify `data-testid="notes-empty"` is no longer rendered
|
|
- [ ] Send "Also remember I live in Berlin."; verify within 15s the notes list grows (previous notes preserved, new note added) — confirms the agent passes the FULL updated list per `set_notes` contract
|
|
|
|
#### UI Writes Back to Agent-Authored Slice (clear notes)
|
|
|
|
- [ ] With notes present, verify `data-testid="notes-clear-button"` is visible
|
|
- [ ] Click the Clear button; verify the notes list disappears and `data-testid="notes-empty"` re-renders
|
|
- [ ] Ask "What do you remember about me?"; verify the agent no longer cites the cleared notes (state was written back by the UI via `agent.setState({ notes: [] })`)
|
|
|
|
#### Multi-Turn State Persistence
|
|
|
|
- [ ] Change tone to `playful` and add the `Music` interest; send "Write me a one-line haiku greeting."; verify the reply is playful and references music
|
|
- [ ] Send a follow-up "Do it again in French."; verify the reply remains playful, switches to French, and still acknowledges the music interest — confirms preferences persist across turns without being re-sent
|
|
- [ ] Reload the page; verify preferences reset to defaults (`tone: casual`, `language: English`, empty interests, empty name) and notes reset to empty (state is per-session, seeded by the page's `useEffect`)
|
|
|
|
### 3. Error Handling
|
|
|
|
- [ ] Attempt to send an empty message; verify it is a no-op (no user bubble, no assistant response)
|
|
- [ ] Deselect all interests and clear the name; send "Who am I?"; verify the agent answers without crashing (the callback skips injection when `preferences` is empty)
|
|
- [ ] Verify DevTools -> Console shows no uncaught errors during any flow above
|
|
|
|
## Expected Results
|
|
|
|
- Page loads within 3 seconds; assistant text response within 10 seconds
|
|
- Preferences writes are reflected in `pref-state-json` synchronously on change
|
|
- Agent-authored notes appear in `notes-card` within 15 seconds of a "remember" prompt, and the full prior list is preserved on subsequent `set_notes` calls
|
|
- Clear button round-trips UI -> agent state and the agent loses access to the cleared notes on the next turn
|
|
- No UI layout breaks, no uncaught console errors
|