operation-tracker

Track reads, edits, and file turns to trigger delegation thresholds.

3|Updated Nov 28, 2025
One-click install
npx skills add https://github.com/rayk/lucid-toolkit --skill operation-tracker
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: operation-tracker
Source: https://github.com/rayk/lucid-toolkit/tree/main/plugins/luc/skills/operation-tracker
Command: npx skills add https://github.com/rayk/lucid-toolkit --skill operation-tracker

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Maintain awareness of operation counts across the conversation to prevent cumulative context exhaustion. Use when you notice multiple operations accumulating or when working on iterative tasks across turns.

Core Features & Use Cases

  • Mental tracking of reads, edits, and file turns
  • Threshold-based delegation to preserve context
  • Patterns for continue loops, read chains, and multi-turn iterations

Quick Start

Before each operation, record reads/edits since last Task; delegate when thresholds are exceeded

Frequently Asked Questions about operation-tracker

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prevent context exhaustion when running multiple operations across conversation turns?

Context exhaustion occurs when cumulative reads, edits, and file operations accumulate across turns. Track operation counts—reads, edits, consecutive same-file turns, and continue cycles—against thresholds (≥2 reads, ≥3 edits, ≥3 consecutive same-file turns, ≥3 continue cycles) to trigger delegation to subagents before context runs out.

When should I delegate iterative tasks to subagents instead of continuing in the same conversation?

Delegate when operation thresholds are exceeded: 2+ reads since last task, 3+ edits, 3+ consecutive turns on the same file, or 3+ continue cycles. Delegation preserves context budget by offloading accumulated work, enabling continuation without exhaustion.

What operations should I count to know when to delegate work in an LLM conversation?

Count file reads, file edits, consecutive turns working on the same file, and continue cycles (conversation restarts). These four metrics signal cumulative context load. When any reaches its threshold, delegation becomes necessary to maintain usable context window.

How does operation tracking help with multi-turn workflows?

Operation tracking maintains awareness of reads, edits, and turn patterns across the conversation. By monitoring these metrics, you identify when accumulated operations degrade context efficiency, allowing proactive delegation before performance degrades or context exhausts.

Can I use operation tracking with any LLM coordination pattern?

Yes. Operation tracking applies to any multi-turn workflow involving delegation. It works with read chains, iterative edits, and continue loops—any pattern where operations accumulate across turns and context preservation matters for task completion.

What's the difference between tracking operations manually versus using thresholds?

Manual tracking requires remembering counts; threshold-based tracking automates the decision. Recording reads, edits, and file turns against fixed thresholds (≥2 reads, ≥3 edits, ≥3 consecutive same-file, ≥3 continues) makes delegation decisions objective and timely.