fix(graph): redact runtime error events

[XnLemon]

Objective

Close the next Phase2 runtime redaction gap by preventing graph/runtime model-visible error surfaces from exposing raw provider, node, tool, or Pregel error details.

Completed Changes

• Added graph-local error redaction helpers that use the platform redactor and fail closed when redaction cannot be initialized.
• Redacted graph node, Pregel, tool execution, model execution, and execution error records before they are exposed through event metadata or model.ResponseError.
• Redacted terminal graph and node runtime error event construction paths.
• Redacted model response error events by emitting a cloned redacted model.Response while preserving the original response for internal callbacks/state.
• Redacted ResponseError.Message, Code, and Param fields for graph/runtime model-visible response errors.
• Added focused tests covering graph error events, execution errors, processModelResponse, and fast model response error emission with sensitive strings.

Validation

• go test ./graph -count=1 -run 'TestNewExecutionError_NilCallbackContext|TestExecutionErrorCollector_CustomResponseError|TestGraphErrorEventsRedactSensitiveMessages|TestNewToolAndModelEvents|TestNewPregelAndChannelStateEvents|TestEmitExecutionError|TestProcessModelResponse_EventAndErrors|TestEmitFastModelResponseEvent'
• go vet ./graph
• git diff --check
• Subagent open-code-review loop completed with 0 High / 0 Medium / 0 Low after fixes.

Known Risks / Limitations

• Full go test ./graph still has pre-existing timing/retry failures unrelated to this redaction diff, as noted during review.
• event.NewErrorEvent remains unchanged globally; this increment redacts graph/runtime callers only.

Follow-up Work

• Continue Phase2 closure on remaining gateway/server model-visible error paths such as OpenClaw gateway/client and server OpenAI converter responses.

[coderabbitai]

Warning

Review limit reached

@XnLemon, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 17 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: 8249000a-2efd-4f0f-b035-efbf0f4dc616

📥 Commits

Reviewing files that changed from the base of the PR and between 361e2a4 and c81e17e.

📒 Files selected for processing (9)

• graph/checkpoint_test.go
• graph/error_redaction.go
• graph/events.go
• graph/events_test.go
• graph/execution_error.go
• graph/execution_error_test.go
• graph/executor.go
• graph/state_graph.go
• graph/state_graph_additional_test.go

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch codex/phase2-runtime-gateway-error-redaction

---

_{Comment @coderabbitai help to get the list of available commands.}