[BenGSchulz]

Summary

Fourth PR of the Wind chat renderer stack (MST-16565): agent tool call messages (AGENTS_TOOL_CALL_RENDERER) and tree messages (APOLLO_CHAT_TREE_RENDERER) render in Wind. Both are registered as Wind built-ins, and a renderer injected on the service still wins (tested).

Tool call: apollo-wind ToolCall

• Status: Material's state maps to the primitive's status.

  • loading → executing;
  • error → failed;
  • success → completed;
  • no state → completed, because Material's line then reads "Ran".

  The line reads "Running 'x'" or "Ran 'x' for 1.23 seconds", as Material's does, and is localized.

• Display modes:

  • ToolNameOnly omits ToolCallContent, so only the status line shows.
  • InputsAndOutputs and FullTrace get the body.
  • The body starts closed, as Material's does, and opens by itself when the trace has escalation task links.

• Sections:

  • Input, Errors and Output use the primitive's value sections. Strings render through CodeBlock (renderValue) and objects through the primitive's JSON tree.
  • FullTrace adds a Traces section: nested trace items with Material's span icons, model calls and agent output left out.
  • Escalation lists the task links with a waiting or completed icon.

• Stable state: which parts are open (the card, Input, Output, Errors, Traces, Escalation and each trace row) is held by the renderer, as in Material, so it survives the finished call being re-sent under the same id and closing and reopening the card. The ported "stays expanded when the tool result replaces the message" test covers it, including an open Input section. Auto-opening for escalation happens once, so a re-send does not reopen a card the reader closed. A failure without input shows as failed. Trace rows and their icons are built level by level, only while the Traces section and the row are open.

Tree: disclosure tree, not apollo-wind TreeView

The tree message renders as a nested disclosure list built from apollo-wind Tooltip and Tailwind. The same component is shared with the tool call's Traces section.

• Rows are open unless their span says expanded: false, as in Material.
• Each row shows the title (with titleColor), a lucide icon via ServiceIcon, and additionalInfo on the right. It gets a tooltip only when the span has a description; a described leaf row is focusable, so keyboard users get the tooltip too. Rows use the tool call section's styling (text-emphasis hover, chevron, indentation), so a trace inside a tool call and a tree message read alike. The tree model no longer supplies a "No description" placeholder, which also removes it from Material.

apollo-wind's TreeView (FileTreeView) is a file-manager component and does not fit here (see the open decision):

• it always shows a search box and English Expand and Collapse buttons;
• it has no initial expanded state;
• it has no description line.

Hoisted into src/chat/headless

Helper	What it carries	Material change
tool-call/tool-call-model.ts	Span types; getToolCallState (from input, output and the span tree, including child errors); getToolCallName; getToolCallDurationSeconds; getTaskUrls; getToolCallValues (input, error and output as Material picks them); getVisibleTraceChildren; getSpanAttributes	ApToolCall.tsx uses all of them
tool-call/span-icon.ts, tool-call/span-icons.ts	The span icon lookup and its SVG set, moved from ap-tool-call/utils and ap-tool-call/assets	ApToolCall.tsx imports them from here
tree/chat-tree.ts: toChatTreeItem, getExpandedTreeIds	The tree row model and default expansion	tree-renderer.tsx default transform and expanded ids
utils/openable-url.ts: isOpenableUrl	The http(s)/relative URL rule from R-3, now shared	ToolCallSection.tsx task links (see below)

Security: escalation task links come from the backend. Both renderers now render them as links only when they are http(s) or relative; anything else shows as plain text. This also changes Material, which previously linked any scheme.

Storybook

pnpm storybook:design, Apollo React / Chat / Playground, then Compare renderers with scenario:

• Tool calls: one call per display mode, then:
  • a call that runs and finishes 3s later under the same id (expand it while it runs);
  • a failed call (Errors);
  • a FullTrace call waiting on two escalation task links, which opens by itself.
• Tree message: the trace as a tree, open by default. Three spans have descriptions (tooltips) and two have none.

Tests

• vitest --run src/chat src/material/components/ap-chat src/material/components/ap-tool-call: 43 files, 237 tests pass. New tests cover:

  • the tool call model: state from input/output and from the span tree, name, duration, task URLs (string attributes, descendants), section values, trace filtering;
  • the tree model: rows, fallbacks, expansion;
  • the Wind tool call:
    • running then finished;
    • stays expanded when the result replaces the message;
    • ToolNameOnly without a toggle;
    • string values as code and objects as a tree;
    • traces only for FullTrace, without model calls;
    • escalation opens by itself and never links a javascript: URL;
    • an injected renderer wins;
  • the Wind tree: open by default, expanded: false closed.

• tsc --noEmit: clean.

• biome check on touched files: no errors; two warnings already present in Material (its trace icon dangerouslySetInnerHTML and an any in ToolCallSection.tsx). The first push had a format error in headless/tree/chat-tree.ts, now fixed.

• New strings, in en.json only:

  • tool call: status names, the running and ran lines, seconds, section titles, the waiting label;
  • JSON tree: the read-only strings (expand, collapse, item and key counts with plurals, copy path and value).

  The running and ran lines double their apostrophes, since a single ' starts a quoted literal in ICU.

Stack

1. R-1 Wind shell: feat(apollo-react): add the Wind chat renderer shell #1371
2. R-2 message list: feat(apollo-react): add the Wind chat message list #1372
3. R-3 markdown renderer: feat(apollo-react): add the Wind chat markdown renderer #1375
4. R-4 tool-call and tree renderers: this PR
5. R-5 composer: feat(apollo-react): add the Wind chat composer #1378
6. R-6 first run, suggestions and thinking state: feat(apollo-react): add the Wind chat first run, suggestions and thinking state #1396
7. R-7 history and settings: feat(apollo-react): add the Wind chat history and settings panels #1397
8. R-8 CSS delivery and docs: feat(apollo-react): document the Wind chat setup and its delivery #1399

Jira: MST-16565

🤖 Generated with Claude Code

[github-actions]

Apollo Coded App preview deployments are ready.

Project	Status	Preview	Updated (PT)
apollo-design	Ready	Preview · Logs	Oct 15, 2026, 02:58:45 PM
apollo-docs	Ready	Preview · Logs	Oct 15, 2026, 02:58:45 PM
apollo-landing	Ready	Preview · Logs	Oct 15, 2026, 02:58:45 PM
apollo-vertex	Ready	Preview · Logs	Oct 15, 2026, 02:58:45 PM

[github-actions]

Dependency License Review

• ✅ 1947 package(s) scanned
• ✅ No license issues found
• ⚠️ 2 package(s) excluded (see details below)

License distribution

License	Packages
MIT	1713
ISC	91
Apache-2.0	56
BSD-3-Clause	28
BSD-2-Clause	23
BlueOak-1.0.0	8
MPL-2.0	4
MIT-0	3
CC0-1.0	3
MIT OR Apache-2.0	2
(MIT OR Apache-2.0)	2
Unlicense	2
LGPL-3.0-or-later	1
Python-2.0	1
CC-BY-4.0	1
(MPL-2.0 OR Apache-2.0)	1
Unknown	1
Artistic-2.0	1
(WTFPL OR MIT)	1
(BSD-2-Clause OR MIT OR Apache-2.0)	1
CC-BY-3.0	1
0BSD	1
(MIT OR CC0-1.0)	1
MIT AND ISC	1

Excluded packages

Package	Version	License	Reason
@img/sharp-libvips-linux-x64	1.3.4	LGPL-3.0-or-later	LGPL pre-built binary, not linked
khroma	2.1.0	Unknown	MIT per GitHub repo, missing license field in package.json

[github-actions]

📊 Coverage + size by package

Per-package coverage and bundle size on this PR. New-line coverage = of the source lines this PR adds or changes, the % hit by tests.

Package	Coverage	New-line coverage	Packed (gzip)	Unpacked	vs main
@uipath/apollo-core	—	—	42.02 MB	50.17 MB	±0
@uipath/apollo-react	56.0%	78.9% (172/218)	8.02 MB	31.57 MB	+32.9 KB
@uipath/apollo-ui-icons	—	—	2.86 MB	6.97 MB	±0
@uipath/apollo-vertex	—	—	—	—	—
@uipath/apollo-wind	—	—	785.2 KB	4.55 MB	+10 B
@uipath/ap-chat	85.8%	—	40.37 MB	45.58 MB	+1.7 KB

_{"Coverage" is each package's own coverage.include scope (e.g. apollo-core instruments only scripts/). "Packed"/"Unpacked" come from npm pack --dry-run and only cover built packages — "—" means not measured this run (package not affected / not built). "vs main" is the packed (gzipped) delta against the last successful main build (the package-sizes artifact from the Release workflow); "—" there means no main baseline was available this run. The baseline is main's latest build, not this PR's exact merge-base, so it includes any drift since the branch diverged. Packages with no vitest config are omitted.}

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.