t1463: prevent cold-start false idle kills in pulse watchdog

[marcusquinn]

Summary

• Add a cold-start watchdog timeout (PULSE_COLD_START_TIMEOUT, default 1200s) so pulse sessions get a longer first-output grace window before stall kill decisions.
• Reorder watchdog checks to evaluate log progress first, then apply CPU idle enforcement only after at least one output/progress event.
• Keep existing stuck-process protection after first progress: standard PULSE_PROGRESS_TIMEOUT and PULSE_IDLE_TIMEOUT continue to guard post-start stalls.

Verification

• shellcheck .agents/scripts/pulse-wrapper.sh
• bash -n .agents/scripts/pulse-wrapper.sh

Closes #4311
Refs #4261

Summary by CodeRabbit

Release Notes

• Bug Fixes
  • Enhanced process watchdog monitoring to more accurately distinguish application startup phases from normal operation, reducing false positive alerts.
  • Improved progress tracking and idle detection with better sequencing to prevent monitoring errors during early execution and ensure more reliable process supervision.

[gemini-code-assist]

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 4ee08e46-d506-4e0a-918f-eb42f8716d50

📥 Commits

Reviewing files that changed from the base of the PR and between 3c953dd and f62b507.

📒 Files selected for processing (1)

• .agents/scripts/pulse-wrapper.sh

---

Walkthrough

The pulse wrapper's watchdog timeout logic is refactored to distinguish between cold-start and post-start phases. A new PULSE_COLD_START_TIMEOUT configuration and has_seen_progress flag enable separate deadlines before initial output. Idle detection now occurs only after the first progress event, preventing false idle kills during startup.

Changes

Cohort / File(s)	Summary
Pulse Watchdog Timing Logic .agents/scripts/pulse-wrapper.sh	Added PULSE_COLD_START_TIMEOUT configuration with validation; introduced has_seen_progress state flag to track first output occurrence; refactored watchdog ordering to apply cold-start timeouts until progress detected, then switch to post-start timeouts; idle CPU-based detection now requires prior progress observation to avoid false positives during startup.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~22 minutes

Possibly related PRs

• t2893: Address quality-debt review feedback on pulse-wrapper.sh #2910: Both PRs refine pulse-wrapper.sh watchdog logic with kill_reason variable handling and watchdog event ordering improvements.
• t1397: fix pulse-wrapper stuck process — add watchdog timeout #2853: Both PRs modify progress timeout handling and idle detection mechanics within the pulse wrapper's watchdog system.
• t1439: fix pulse status drift and worker cap reporting #4140: Both PRs update pulse-wrapper.sh initialization and configuration, adding new timeout settings and runtime validation logic.

Suggested labels

bug

Poem

🕐 Cold starts no longer stumble in the dark,
A patient watchdog now waits for the spark,
Progress unlocks the idle-check gate,
No false alarms before processes wake—
Two-phase timing keeps jobs running straight! ✨

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and specifically describes the main change: preventing cold-start false idle kills in the pulse watchdog, which directly matches the PR's primary objective.
Linked Issues check	✅ Passed	The PR successfully implements the core requirement from #4311: differentiating cold-start inactivity from genuine stalls by introducing PULSE_COLD_START_TIMEOUT and reordering watchdog checks to run post-first-output.
Out of Scope Changes check	✅ Passed	All changes are directly related to the cold-start watchdog timeout feature; no unrelated modifications detected outside the scope of fixing false idle kills.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

• 📝 Generate docstrings (stacked PR)
• 📝 Generate docstrings (commit on current branch)

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch bugfix/pulse-idle-watchdog-autonomy

📝 Coding Plan

• Generate coding plan for human review comments

---

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.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[github-actions]

🔍 Code Quality Report

�[0;35m[MONITOR]�[0m Code Review Monitoring Report

�[0;34m[INFO]�[0m Latest Quality Status:
SonarCloud: 0 bugs, 0 vulnerabilities, 407 code smells

�[0;34m[INFO]�[0m Recent monitoring activity:
Fri Mar 13 01:17:01 UTC 2026: Code review monitoring started
Fri Mar 13 01:17:02 UTC 2026: SonarCloud - Bugs: 0, Vulnerabilities: 0, Code Smells: 407

📈 Current Quality Metrics

• BUGS: 0
• CODE SMELLS: 407
• VULNERABILITIES: 0

Generated on: Fri Mar 13 01:17:04 UTC 2026

---

Generated by AI DevOps Framework Code Review Monitoring

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud