Summary

Fixes #10867 — session_status tool was not passing session-level thinking/verbose/reasoning/elevated overrides to buildStatusMessage, causing it to fall back to agent defaults (e.g. "Think: off") instead of the actual session value.

Root Cause

The /status command path (commands-status.ts) correctly passes resolvedThink, resolvedVerbose, resolvedReasoning, and resolvedElevated from the session entry to buildStatusMessage. However, the session_status tool (session-status-tool.ts) did not pass these parameters, so buildStatusMessage fell back to agent.thinkingDefault ?? "off".

Changes

• src/agents/tools/session-status-tool.ts: Pass resolvedThink, resolvedVerbose, resolvedReasoning, resolvedElevated from resolved.entry to buildStatusMessage, matching the /status command path.
• src/agents/openclaw-tools.session-status.test.ts: Add regression test verifying thinkingLevel: "medium" displays as Think: medium (not Think: off).
• CHANGELOG.md: Add fix entry.

Testing

pnpm vitest run src/agents/openclaw-tools.session-status.test.ts --reporter=verbose
# 8/8 passed (including new regression test)
pnpm build  # ✅
pnpm oxlint  # 0 warnings, 0 errors

Greptile Overview

Greptile Summary

• Fixes session_status tool output by passing session-level thinking/verbose/reasoning/elevated overrides into buildStatusMessage, matching the /status command behavior.
• Adds a regression test asserting a stored thinkingLevel: "medium" renders as Think: medium (and not the agent default).
• Updates CHANGELOG with an entry for the fix.

Confidence Score: 5/5

• This PR is safe to merge with minimal risk.
• Changes are small and localized: they pass through already-existing resolved override fields to buildStatusMessage, matching the established /status code path, and include a focused regression test covering the previously incorrect behavior.
• No files require special attention