session-health

Evaluate conversation health and recommend continue, compact, checkpoint, or fresh-session actions.

Updated Feb 20, 2026
One-click install
npx skills add https://github.com/supertyrelle/pelley --skill session-health-supertyrelle
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: session-health
Source: https://github.com/supertyrelle/pelley/tree/main/skills/session-health
Command: npx skills add https://github.com/supertyrelle/pelley --skill session-health-supertyrelle

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you assess whether a long-running conversation is still healthy, so you can catch context overload, scope drift, and declining answer quality before they affect the work.

Core Features & Use Cases

  • Context Load Assessment: Estimates how much of the session has been consumed by files read, tasks dispatched, and topic breadth.
  • Scope Coherence Review: Compares the original goal with current work to detect drift and decide whether to continue, checkpoint, or reset.
  • Quality Check: Flags signs of repetition, weak assumptions, broadening searches, and lost task state.
  • Session State Snapshot: Reviews in-progress work, available tasks, and uncommitted changes to guide the next move.

Quick Start

Ask the Skill to evaluate the current conversation health and recommend whether to continue, compact, checkpoint, or start a fresh session.

Frequently Asked Questions about session-health

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

FAQPage Schema
How do I check for context drift in a long-running coding session?

Evaluate session health by reviewing workload estimation and scope coherence to detect context drift. This diagnostic compares original goals against current work to flag scope drift and context overload before they affect progress.

What is the best way to diagnose conversation health before a major task transition?

The best way to diagnose conversation health is to perform a session-state review. This assesses available tasks, in-progress work, and uncommitted changes to provide a clear recommendation for your next move.

When should I compact or checkpoint an agentic coding session?

Compact or checkpoint an agentic coding session when a health diagnostic flags repetition, weak assumptions, or broadening searches. Evaluating session state determines whether context load requires compaction, checkpointing, or starting a fresh session.

Why does my AI coding assistant lose task state and repeat itself in long sessions?

Your AI coding assistant loses task state due to context overload and scope drift in long sessions. Running a session health diagnostic identifies these quality issues and recommends whether to continue, compact, or reset.

Can I assess context load and scope coherence without external dependencies?

Yes, you can assess context load and scope coherence without external dependencies. The evaluation relies entirely on session-state review, workload estimation, and scope checks to recommend continue, compact, or fresh-session actions.