mirror of
https://github.com/CopilotKit/CopilotKit.git
synced 2026-09-14 16:26:20 +08:00
Jordan Ritter
5f21584b3a
Refresh d6/langgraph-python fixtures (beautiful-chat, chat-slots,
frontend-tools-async, gen-ui-{agent,custom,declarative,interrupt,open},
headless-complete, hitl-in-app, interrupt-headless, shared-state-streaming,
tool-rendering, tool-rendering-reasoning-chain, _from-feature-parity) plus
d4/langgraph-python/chat.json. Combined with the conveyance shim landed
in this branch, LGP is fixture-complete at 180 pass / 5 fail / 2 skip.
The residual 5 fails trace to react-core/v2's custom-wildcard-renderer
bug surfaced through gen-ui-interrupt (shared useInterrupt hook misbehaves
on the second interrupt; cross-integration D6 blocker, not a fixture or
conveyance defect). The 2 skips are the by-design mcp-apps iframe gap.
41 lines
2.3 KiB
JSON
41 lines
2.3 KiB
JSON
{
|
|
"_meta": {
|
|
"description": "D6 fixtures for langgraph-python / gen-ui-custom (pie chart path)",
|
|
"sourceFile": "harness/fixtures/d5/gen-ui-custom.json",
|
|
"created": "2026-05-21"
|
|
},
|
|
"fixtures": [
|
|
{
|
|
"_comment": "gen-ui-custom pie chart — first leg: emit render_pie_chart tool call. hasToolResult:false ensures this only matches when there's no prior tool result in the conversation. Includes narration content (with 'pie chart' tokens) inline so the D5 probe's last-assistant-text assertion passes even when the LGP runtime does not re-invoke the LLM after a frontend tool result returns (the second-leg fixture below remains a safety net for runtimes that DO loop back). chunkSize: 9999 guards against the @ag-ui/langgraph 0.0.34 partial-tool-call JSON.parse class — the render_pie_chart args are large enough to otherwise stream across multiple chunks and trip RUN_ERROR.",
|
|
"match": {
|
|
"userMessage": "Show me a pie chart of revenue by category",
|
|
"hasToolResult": false,
|
|
"context": "langgraph-python"
|
|
},
|
|
"response": {
|
|
"content": "Pie chart rendered above — Electronics is the largest slice, followed by Clothing, Food, and Books.",
|
|
"toolCalls": [
|
|
{
|
|
"id": "call_d5_render_pie_chart_001",
|
|
"name": "render_pie_chart",
|
|
"arguments": "{\"title\":\"Revenue by Category\",\"description\":\"Revenue breakdown by product category (Q4)\",\"data\":[{\"label\":\"Electronics\",\"value\":42000},{\"label\":\"Clothing\",\"value\":28000},{\"label\":\"Food\",\"value\":18000},{\"label\":\"Books\",\"value\":12000}]}"
|
|
}
|
|
]
|
|
},
|
|
"chunkSize": 9999
|
|
},
|
|
{
|
|
"_comment": "gen-ui-custom pie chart — second leg: narration after the pie chart tool result lands. Kept as a safety net when the LGP runtime DOES loop back to the LLM. The follow-up text contains 'pie' and 'chart' tokens — the D5 probe asserts their presence.",
|
|
"match": {
|
|
"userMessage": "Show me a pie chart of revenue by category",
|
|
"hasToolResult": true,
|
|
"context": "langgraph-python"
|
|
},
|
|
"response": {
|
|
"content": "Pie chart rendered above — Electronics is the largest slice, followed by Clothing, Food, and Books."
|
|
},
|
|
"chunkSize": 9999
|
|
}
|
|
]
|
|
}
|