chore(deps): bump temporalio from 1.24.0 to 1.29.0

[dependabot]

Bumps temporalio from 1.24.0 to 1.29.0.

Release notes

Sourced from temporalio's releases.

1.29.0

Notable Changes

Added

• Added experimental temporalio.workflow.signal_with_start_workflow, backed by generated system Nexus bindings for WorkflowService.SignalWithStartWorkflowExecution.
• Added OpenAI Agents plugin support for CustomTool dispatch, including lazy tool discovery through defer_loading.

Changed

• Client connections now use gzip transport-level gRPC compression by default. Pass grpc_compression=GrpcCompression.NONE to Client.connect or CloudOperationsClient.connect to disable it.

Breaking Changes

• StartWorkflowUpdateWithStartInput now owns the authoritative rpc_metadata and rpc_timeout fields for OutboundInterceptor.start_update_with_start_workflow. These fields were removed from the nested update-with-start input objects, so custom interceptors that accessed them there should read or update the top-level fields instead.

Fixed

• Fixed breakpoint() and pdb.set_trace() inside workflow code when a worker runs with debug_mode=True or TEMPORAL_DEBUG=1; sandboxed workflows without debug mode now get a clearer error pointing to debug_mode=True.

• Fixed start_update_with_start_workflow interceptor handling so RPC metadata and timeouts are forwarded to the underlying execute_multi_operation call.

• Fixed OpenAI Agents plugin streamed event serialization when pydantic had not yet built deferred schemas, and fixed terminal sandbox errors retrying forever.

• Removed the lazy-connect lock from the per-RPC hot path. It was previously acquired on every RPC, putting an event-loop-bound primitive on the hot path; it is now skipped once the client is connected. This reduces the client's coupling to the event loop it connected on, which can help when reusing a single long-lived Client across event loops or threads (e.g. the dedicated-loop pattern used with gevent/gunicorn and synchronous services). Note this does not make a Client fully thread- or loop-agnostic; reusing one long-lived loop is still the recommended pattern.

  SDK Core

  • 993de3fb deps: sync api_upstream subtree to temporal-api@6c8bd13 (#1293)
  • d0065e38 Rename master to main for API deps (#1309)
  • f57f6f06 fix(sdk): no longer expose raw proto for application error category (#1307)

... (truncated)

Changelog

Sourced from temporalio's changelog.

[1.29.0] - 2026-06-17

Added

• Added experimental temporalio.workflow.signal_with_start_workflow, backed by generated system Nexus bindings for WorkflowService.SignalWithStartWorkflowExecution.
• Added OpenAI Agents plugin support for CustomTool dispatch, including lazy tool discovery through defer_loading.

Changed

• Client connections now use gzip transport-level gRPC compression by default. Pass grpc_compression=GrpcCompression.NONE to Client.connect or CloudOperationsClient.connect to disable it.

Breaking Changes

• StartWorkflowUpdateWithStartInput now owns the authoritative rpc_metadata and rpc_timeout fields for OutboundInterceptor.start_update_with_start_workflow. These fields were removed from the nested update-with-start input objects, so custom interceptors that accessed them there should read or update the top-level fields instead.

Fixed

• Fixed breakpoint() and pdb.set_trace() inside workflow code when a worker runs with debug_mode=True or TEMPORAL_DEBUG=1; sandboxed workflows without debug mode now get a clearer error pointing to debug_mode=True.
• Fixed start_update_with_start_workflow interceptor handling so RPC metadata and timeouts are forwarded to the underlying execute_multi_operation call.
• Fixed OpenAI Agents plugin streamed event serialization when pydantic had not yet built deferred schemas, and fixed terminal sandbox errors retrying forever.
• Removed the lazy-connect lock from the per-RPC hot path. It was previously acquired on every RPC, putting an event-loop-bound primitive on the hot path; it is now skipped once the client is connected. This reduces the client's coupling to the event loop it connected on, which can help when reusing a single long-lived Client across event loops or threads (e.g. the dedicated-loop pattern used with gevent/gunicorn and synchronous services). Note this does not make a Client fully thread- or loop-agnostic; reusing one long-lived loop is still the recommended pattern.

Commits

• 4b4e34e Bump version to 1.29.0 (#1607)
• 530777e Fix asyncio lock contention in client calls (#1606)
• 3c211b3 Translate terminal sandbox errors to non-retryable ApplicationError (#1595)
• 6886d34 Upgrade PyO3 to 0.29 (#1599)
• 29dc19c Fix pdb / breakpoint() hang in workflow code (#1104) (#1568)
• 7ab0e9d Gzip compression options (#1587)
• 0cf8b3e System Nexus payload handling from WIT (#1572)
• 53ae9fc Fix error message for tracer provider initialization (#1584)
• b9b2cc3 Fix interceptor contract inconsistency for start_update_with_start_workflow (...
• 24badcf AI-249: Support CustomTool in OpenAI Agents plugin tool dispatch (#1570)
• Additional commits viewable in compare view

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

---

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

• @dependabot rebase will rebase this PR
• @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)