imabusyman/loop-operator
prompts/squad/skills/loop-operator/SKILL.md
Supervise an autonomous multi-agent loop, detecting stalls, managing timeouts, and performing safe interventions when agents are blocked or slow. Monitor agent health, log decisions, and escalate issues. Use this in the Orchestrator parallel path to ensure the autonomous loop makes progress and doesn't deadlock or hang.
testing
Updated Apr 25, 2026
$ install --global
skillsauthnpx skillsauth add imabusyman/CodebrewRouter loop-operatorInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 25, 2026, 7:21 PM147.8s1 file scanned
SKILL.md
- name:
- loop-operator
- description:
- Supervise an autonomous multi-agent loop, detecting stalls, managing timeouts, and performing safe interventions when agents are blocked or slow. Monitor agent health, log decisions, and escalate issues. Use this in the Orchestrator parallel path to ensure the autonomous loop makes progress and doesn't deadlock or hang.
Loop Operator — autonomous loop supervision and recovery
A guardian process that monitors the Orchestrator's autonomous loop for stalls, timeouts, and deadlocks. Detects problems and intervenes safely.
When to use
- Orchestrator is running autonomously (no human
/agent squadgate). - You want health monitoring and automatic recovery from stalls.
- Agent tasks may hang, timeout, or deadlock; intervention required.
Responsibilities
1. Health monitoring
Periodically check:
- Task queue progress: Is the next task moving forward, or stuck in
[in_progress]? - Agent responsiveness: Has the agent emitted a status update in the last N minutes?
- Resource usage: Is memory/CPU growing unbounded?
2. Detection rules
Stall detected when:
- A task has been
[in_progress]for >task_timeout(e.g., 1 hour). - No
[CHECKPOINT]or progress message from agent in > 30 minutes. - Task queue is empty but end-to-end progress is incomplete.
Deadlock detected when:
- Multiple tasks are mutually blocked on each other's completion.
- A task depends on a predecessor, but the predecessor is stalled.
3. Interventions
Safe restart
If a task stalls:
- Warn: Log
[BLOCKED]with reason. - Retry: Reassign the task to a fresh agent instance (same role).
- Log: Record the retry in reasoning log with timestamp + reason.
Escalation
If a task fails twice:
- Mark as
[blocked]: Remove from queue. - Notify Conductor: Log the issue and human decision is needed.
- Halt loop: Pause other tasks; wait for human override.
Timeout enforcement
If total session > session_timeout (e.g., 8 hours):
- Checkpoint: Save all partial work.
- Emit report: Summarize progress + incomplete tasks.
- Graceful shutdown: Kill remaining agents; preserve state for recovery.
4. Recovery from restart
When restarting an interrupted loop:
- Read reasoning log from
Docs/squad/runs/<ts>/reasoning.log.md. - Find last checkpoint: Identify which tasks completed.
- Resume queue: Start with the next incomplete task.
- Revalidate file locks: Ensure no two agents claim the same file.
Configuration
loop_operator:
task_timeout: 1h
checkpoint_interval: 30m
stall_detection_interval: 5m
max_retries: 2
session_timeout: 8h
escalation_notify: "conductor" # or "user"
heartbeat_required: true
Heartbeat protocol
Each agent must emit periodic heartbeats:
[CHECKPOINT] Task X, step 5/10, 50% done. ETA 20 mins.
Loop operator reads handoff artifact's progress field; if no update for 30 mins → stall detected.
Decisions logged
## 2026-04-20T16:00:00Z — loop-operator — MEDIUM
Decision: Task 1 (Coder) timed out after 1h. Reassigned to fresh Coder.
Rationale: No heartbeat for 45 mins; task was hanging on API call.
Evidence: .worktrees/task-1/logs/coder.log shows [CHECKPOINT] stopped at t=0:15.
Action: Killed agent; reset worktree; re-dispatched with timeout=30m.
See also
autonomous-agent-harness— task queue driver (Loop Operator monitors it).subagent-driven-development— defines agent structure that Loop Operator supervises.
Related Skills
imabusyman/write-a-prd
development
VerifiedTrustedCommunity
Generate a Product Requirements Document from a task description, user intent, or PRD outline. Structure the PRD with sections: Overview, Problem Statement, Goals, Scope, Features, Acceptance Criteria, Metrics. Use this when the Planner or Orchestrator needs to formalize requirements before decomposing into implementation steps.
SKILL.mdUpdated Apr 25, 2026
imabusyman/team-builder
development
VerifiedTrustedCommunity
Compose parallel teams of specialized agents for a multi-faceted task. Map subtasks to agent roles, balance workload, and validate team composition. Use this when breaking a task into parallel streams that require different expertise (Coder, Tester, Infra, etc.).
SKILL.mdUpdated Apr 25, 2026
imabusyman/subagent-driven-development
development
VerifiedTrustedCommunity
Delegate implementation work to specialized subagent instances, each inheriting a task scope and constraints from the parent. Emit structured handoffs, monitor completion signals, aggregate results. Use this in the Orchestrator parallel path when spawning independent Coder/Tester/Infra subagents from a single coordinator.
SKILL.mdUpdated Apr 25, 2026
imabusyman/reasoning
testing
VerifiedTrustedCommunity
Maintain the append-only reasoning log and write handoff envelopes for every squad delegation. Use this when the Conductor delegates to a specialist or when a specialist records a non-trivial decision.
SKILL.mdUpdated Apr 25, 2026