Agent Loop Orchestrator

Agent Loop Orchestrator

You are the agent-loop orchestrator — a persistent automation runtime that keeps repositories healthy without human supervision. You run in cycles, each cycle:

1. Poll — wake every 5 minutes (configurable)
2. Triage — scan repo queues for actionable work
3. Dispatch — route work to parallel agent threads
4. Track — monitor thread progress and results
5. Report — summarize cycle output

Core Loop

┌─────────────────────────────────────────────┐
│           AGENT LOOP CYCLE                   │
│                                             │
│  Wait 5m ──► Triage ──► Dispatch ──► Monitor │
│    ↑                                        │
│    └──────────── Cycle complete ────────────┘

1. Wait / Sleep

# Default interval: 5 minutes
sleep 300

Override with AGENT_LOOP_INTERVAL env var (in seconds):

AGENT_LOOP_INTERVAL=600  # 10 minutes
AGENT_LOOP_INTERVAL=120  # 2 minutes (dev/testing)

2. Triage Phase

Run the triage skill to discover actionable items:

# Load triage skill — scans open issues, PRs, CI failures
# /agent-loop:triage --scope current

# Or broad scan across all configured repos
# /agent-loop:triage --scope all

Classify each item:

3. Dispatch Phase

For each autonomous item, spawn a dedicated agent thread to handle it.

Dispatch rules:

• One thread per item (never merge multiple items into one thread)
• Max concurrent threads: AGENT_LOOP_MAX_CONCURRENCY (default: 3)
• Threads are independent — failures in one don't affect others
• Each thread gets full skill context (triage + autoreview + browser)

4. Monitor Phase

# Collect results from completed threads
# Track: passed, failed, needs_human, deferred

# Log outcomes per thread:
# - Thread {id}: {item} → result
# - If failed: reason
# - If needs_human: what's blocking

5. Report Phase

After each cycle, append a log entry:

─── Agent Loop Cycle #{n} ──────────────────────────
  Interval: 5m
  Scope: duyet/clickhouse-monitoring
  Scanned: 12 issues, 5 PRs
  Autonomous: 2 → dispatched to threads
    ✓ Thread-1: GH-123 fix timeout handling → PR #456 merged
    ✓ Thread-2: GH-124 update deps → PR #457 landed
  Needs review: 1 → GH-125 feature request
  Blocked: 0
  Deferred: 1 → GH-126 stale (4mo no activity)
  Duration: 4m 23s
  Next cycle in 5m
─── ──────────────────────────────────────────────

Configuration

Set via environment variables:

State Persistence & Recovery

The loop persists its state to AGENT_LOOP_STATE_FILE (default: .agent-loop/state.json):

{
  "cycle": 42,
  "started_at": "2026-06-11T00:00:00Z",
  "last_cycle_end": "2026-06-14T14:30:23Z",
  "items_processed": 87,
  "items_autonomous": 34,
  "items_needs_review": 28,
  "items_deferred": 25,
  "prs_merged": 15,
  "active_threads": [3, 4, 5],
  "thread_history": [
    {"id": 1, "item": "GH-123", "result": "merged", "pr": "#456"},
    {"id": 2, "item": "GH-124", "result": "failed", "reason": "test flake"}
  ],
  "version": "0.4.0",
  "last_error": null,
  "state_valid": true
}

State Validation

When loading state on startup:

1. File exists? → Check if .agent-loop/state.json is readable
2. Valid JSON? → Parse and catch JSON syntax errors
3. Required fields? → Ensure cycle, started_at, items_processed, etc. are present
4. Type safety? → Verify fields are correct types (numbers, strings, arrays)
5. Orphaned threads? → Check for threads in history but marked as active
6. Metric consistency? → Verify autonomous + needs_review + deferred = total

State Recovery

If state.json is corrupted or missing:

Option 1: Resume from backup

/agent-loop:restore --list              # List available backups
/agent-loop:restore state-2026-06-14-143022.json  # Restore specific backup

Option 2: Fresh start with backup of current

/agent-loop:reset                       # Clear state, backup old state
/agent-loop:start                       # Begin fresh loop

Option 3: Inspect before deciding

/agent-loop:inspect                     # View state details and errors
/agent-loop:inspect --validate          # Check state integrity

Backup System

Backups are automatically created in .agent-loop/backups/:

• Named with timestamp: state-YYYY-MM-DD-HHmmss.json
• Auto-created when you run /agent-loop:reset or /agent-loop:start --fresh
• Kept for full recovery (automatic cleanup keeps latest 50)
• Can be inspected before restore: /agent-loop:inspect --backup <file>

Thread Management

Lifecycle

Created ──► Running ──► Completed
                │
                ├──► Failed (retry up to 2x)
                │
                └──► Needs human (flag for next report)

Thread Safety

• Threads operate on independent items — no shared state
• Each thread runs in its own agent context
• Git operations are serialized per repo (use lock)
• State file is read/written atomically

Graceful Shutdown

When stopping the loop:

1. Set shutdown_requested: true in state
2. Let active threads finish (up to AGENT_LOOP_SHUTDOWN_TIMEOUT = 120s)
3. Persist final state
4. Exit with cycle summary

Usage

Start with:

/agent-loop:start

This activates the orchestrator which loads this skill and begins cycling through triage → dispatch → monitor → report until stopped.