Uh oh!
There was an error while loading. Please reload this page.
fix(ai-persistence): preserve cancelled tool resumes - #1090
Merged
AlemTuzlak merged 3 commits intoAug 12, 2026
Conversation
Contributor
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (1)
📝 WalkthroughWalkthroughPersistence resume-state reconstruction now includes cancelled tool-call IDs. Cancellation tests verify approval and client-tool handling, including persisted client-tool cancellation, completion, and clearing of pending interrupts. ChangesCancelled resume state
Estimated code review effort: 2 (Simple) | ~10 minutes Sequence Diagram(s)sequenceDiagram
participant PersistedInterrupts
participant PersistenceMiddleware
participant ToolExecutor
PersistedInterrupts->>PersistenceMiddleware: provide cancelled interrupt entries
PersistenceMiddleware->>PersistenceMiddleware: collect cancelledToolCallIds
PersistenceMiddleware->>ToolExecutor: return cancellation-only resumeToolState
ToolExecutor->>ToolExecutor: emit cancellation error result
Possibly related PRs
Suggested labels: Suggested reviewers: 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ 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 |
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@packages/ai-persistence/tests/interrupts.test.ts`:
- Around line 381-397: Update the client-tool fixtures used by the test around
persistClientToolTurn and chat to use toolDefinition() with shared metadata
defined by a Zod schema such as z.object({ query: z.string() }). Keep the
fixture client-only by leaving it without execute or marking it with .client(),
and do not use .server().
🪄 Autofix
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro Plus
Run ID: 45f26a9a-92b8-46bf-aeed-37b61331b835
📒 Files selected for processing (1)
packages/ai-persistence/tests/interrupts.test.ts
Comment on lines
+381
to
+397
| it('completes a cancelled client-tool resume from persisted state with empty client messages', async () => { | ||
| const persistence = memoryPersistence() | ||
| await persistClientToolTurn(persistence, [clientTool('clientSearch')]) | ||
| const pending = await persistence.stores.interrupts!.get( | ||
| 'client_tool_tool-call-1', | ||
| ) | ||
| expect(pending?.status).toBe('pending') | ||
| const afterCancel = mockAdapter([ | ||
| [runStarted(), text('cancelled-and-done'), runFinished('r1')], | ||
| ]) | ||
| const chunks = await collect( | ||
| chat({ | ||
| adapter: afterCancel.adapter, | ||
| messages: [], | ||
| tools: [clientTool('clientSearch')], |
Contributor
There was a problem hiding this comment.
📐 Maintainability & Code Quality | 🟠 Major | ⚡ Quick win
🧩 Analysis chain
🏁 Script executed:
#!/bin/bashset -euo pipefail
# Find repository examples of the required tool builder and implementations.
rg -n -C 4 --glob '*.{ts,tsx}''\btoolDefinition\s*\(' packages
rg -n -C 3 --glob '*.{ts,tsx}''\.(server|client)\s*\(' packagesRepository: TanStack/ai
Length of output: 50368
🏁 Script executed:
#!/bin/bashset -euo pipefail
echo'--- target file imports and fixture definitions ---'
sed -n '1,130p' packages/ai-persistence/tests/interrupts.test.ts
rg -n -C 8 'clientTool|persistClientToolTurn|toolDefinition|from .*zod' packages/ai-persistence/tests/interrupts.test.ts
echo'--- toolDefinition API declarations and focused examples ---'
rg -n -C 6 'export .*toolDefinition|function toolDefinition|class ToolDefinition|\.server\(|\.client\(' packages/ai-core packages/ai-client packages/ai-persistence --glob '*.{ts,tsx}'| head -n 240Repository: TanStack/ai
Length of output: 26750
🏁 Script executed:
#!/bin/bashset -euo pipefail
echo'--- candidate source files ---'
fd -i 'tool' packages --type f | grep -E '(definition|types|index)\.(ts|tsx)$'| head -n 120
echo'--- toolDefinition declarations and Tool types ---'
rg -n -C 12 'toolDefinition|export type Tool|interface Tool|type ToolDefinition' packages --glob '*.{ts,tsx}' \
| grep -v '/tests/'| head -n 320
echo'--- persistence test imports and all fixture uses ---'
rg -n -C 3 'from .*(ai|zod)|clientTool\(|approvalClientTool\(' packages/ai-persistence/tests/interrupts.test.tsRepository: TanStack/ai
Length of output: 24031
🏁 Script executed:
#!/bin/bashset -euo pipefail
echo'--- tool-definition implementation ---'
wc -l packages/ai/src/activities/chat/tools/tool-definition.ts
cat -n packages/ai/src/activities/chat/tools/tool-definition.ts
echo'--- exports and core tool types ---'
rg -n -C 8 'tool-definition|ToolDefinition|ServerTool|ClientTool|interface Tool|type Tool' packages/ai/src packages/ai-client/src --glob '*.{ts,tsx}'| head -n 360
echo'--- nearby persistence test tool fixtures and imports ---'
sed -n '1,145p' packages/ai-persistence/tests/interrupts.test.tsRepository: TanStack/ai
Length of output: 45004
🏁 Script executed:
#!/bin/bashset -euo pipefail
echo'--- runtime classification of tools without execute ---'
rg -n -C 10 'execute\?|typeof .*execute|__toolSide|ClientToolRequest|client tool|client-tool' packages/ai/src/activities packages/ai/src --glob '*.{ts,tsx}'| head -n 360
echo'--- target test assertions for the client-tool fixtures ---'
sed -n '260,455p' packages/ai-persistence/tests/interrupts.test.ts
echo'--- imports and package test conventions for Zod/toolDefinition ---'
rg -n -C 5 'from .*(zod|`@tanstack/ai`)|toolDefinition\(\{' packages/ai-persistence/tests packages/ai-persistence/src --glob '*.{ts,tsx}'| head -n 220Repository: TanStack/ai
Length of output: 50367
🏁 Script executed:
#!/bin/bashset -euo pipefail
echo'--- exact tool execution branch ---'
sed -n '760,900p' packages/ai/src/activities/chat/tools/tool-calls.ts
echo'--- Tool base type and schema conversion ---'
rg -n -C 14 'export interface Tool|export type Tool|type AnyTool|inputSchema' packages/ai/src/types.ts packages/ai/src --glob '*.ts' \
| head -n 260
echo'--- direct-definition documentation and client-side examples ---'
sed -n '276,330p' packages/ai/src/activities/chat/tools/tool-definition.ts
rg -n -C 5 'toolDefinition\(\{[^}]*name: .*(client|lookup|search)|\.client\(\)' packages/ai-client/tests packages/ai/src --glob '*.{ts,tsx}'| head -n 180Repository: TanStack/ai
Length of output: 31148
🏁 Script executed:
#!/bin/bashset -euo pipefail
python3 - <<'PY'from pathlib import Pathimport retest = Path("packages/ai-persistence/tests/interrupts.test.ts").read_text()definition = Path("packages/ai/src/activities/chat/tools/tool-definition.ts").read_text()calls = Path("packages/ai/src/activities/chat/tools/tool-calls.ts").read_text()raw = re.search( r"const clientTool = \(name: string\): Tool => \(\{(?P<body>.*?)\n\}\)", test, re.S,)assert raw, "clientTool fixture not found"body = raw.group("body")assert "execute" not in bodyassert "inputSchema" not in bodyassert "toolDefinition" not in bodyassert "Used directly in chat() on the server" in definitionassert re.search(r"server:\s*<", definition)assert re.search(r"client:\s*<", definition)client_branch = re.search( r"// CASE 1: Client-side tool \(no execute function\)(?P<body>.*?)(?=\n\s*// CASE 2:|\Z)", calls, re.S,)assert client_branch, "client-side execution branch not found"assert "if (!tool.execute)" in callsassert "if (tool.inputSchema && isStandardSchema(tool.inputSchema))" in callsprint("raw_fixture_has_execute:", "execute" in body)print("raw_fixture_has_input_schema:", "inputSchema" in body)print("tool_definition_supports_direct_server_use:", True)print("tool_definition_exposes_server_and_client_builders:", True)print("runtime_classifies_no_execute_as_client_tool:", True)print("runtime_validates_standard_input_schema:", True)PYRepository: TanStack/ai
Length of output: 419
Use toolDefinition() for the client-tool fixtures.
Define the shared metadata with a Zod schema, such as z.object({ query: z.string() }). Keep this client-only fixture as a bare definition or use .client(); .server() is not required because chat() treats tools without execute as client-side.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@packages/ai-persistence/tests/interrupts.test.ts` around lines 381 - 397,
Update the client-tool fixtures used by the test around persistClientToolTurn
and chat to use toolDefinition() with shared metadata defined by a Zod schema
such as z.object({ query: z.string() }). Keep the fixture client-only by leaving
it without execute or marking it with .client(), and do not use .server().
Source: Coding guidelines
View your CI Pipeline Execution ↗ for commit 151423d
☁️ Nx Cloud last updated this comment at |
@tanstack/ai@tanstack/ai-acp@tanstack/ai-angular@tanstack/ai-anthropic@tanstack/ai-bedrock@tanstack/ai-byteplus@tanstack/ai-claude-code@tanstack/ai-client@tanstack/ai-code-mode@tanstack/ai-code-mode-skills@tanstack/ai-codex@tanstack/ai-cohere@tanstack/ai-devtools-core@tanstack/ai-durable-stream@tanstack/ai-elevenlabs@tanstack/ai-event-client@tanstack/ai-fal@tanstack/ai-gemini@tanstack/ai-grok@tanstack/ai-grok-build@tanstack/ai-groq@tanstack/ai-isolate-cloudflare@tanstack/ai-isolate-daytona@tanstack/ai-isolate-node@tanstack/ai-isolate-quickjs@tanstack/ai-isolate-quickjs-bun@tanstack/ai-mcp@tanstack/ai-memory@tanstack/ai-mistral@tanstack/ai-ollama@tanstack/ai-openai@tanstack/ai-opencode@tanstack/ai-openrouter@tanstack/ai-persistence@tanstack/ai-preact@tanstack/ai-react@tanstack/ai-react-ui@tanstack/ai-sandbox@tanstack/ai-sandbox-cloudflare@tanstack/ai-sandbox-daytona@tanstack/ai-sandbox-docker@tanstack/ai-sandbox-local-process@tanstack/ai-sandbox-sprites@tanstack/ai-sandbox-vercel@tanstack/ai-solid@tanstack/ai-solid-ui@tanstack/ai-svelte@tanstack/ai-utils@tanstack/ai-vue@tanstack/ai-vue-ui@tanstack/openai-base@tanstack/preact-ai-devtools@tanstack/react-ai-devtools@tanstack/solid-ai-devtoolscommit: |
AlemTuzlak
approved these changes
Aug 12, 2026
| @@ -829,6 +844,7 @@ describe('interrupt persistence', () => { | |||
| run.calls[0] as { clientToolResults?: ReadonlyMap<string, unknown> } | |||
| ).clientToolResults | |||
Contributor
There was a problem hiding this comment.
[suggestion] The new assertions only prove the translator wrote cancelledToolCallIds. Both cancel tests still use empty messages, no tools, and no stored thread with a pending assistant tool call. The engine therefore never reaches checkForPendingToolCalls() / executeToolCalls(), so these tests cannot fail if a cancelled hydrated client tool still re-interrupts. The file already has the right shape for that path in applies persisted approval and client-tool resume decisions with empty client messages (persist tool-call turn, resume with empty client messages, assert chunks).
Suggestion: Add a regression that mirrors that hydrate path for status: 'cancelled': persist a client-tool interrupt plus the assistant tool-call message, resume with empty messages and the same tools, and assert a TOOL_CALL_RESULT of Tool execution cancelled plus no new client_tool_* interrupt. Keep the current onConfig Set check if you want, but do not treat it as coverage for #1088.
Uh oh!
There was an error while loading. Please reload this page.
Merged
This was referenced Aug 13, 2026
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 freeto 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.
Summary
Fixes#1088
Validation
The repository Playwright/Chromium install was blocked during Windows extraction, so the E2E suite was not run.
Summary by CodeRabbit