context-surfing

Name: context-surfing
Availability: InStock
Author: pskoett

Monitor AI context window health and detect drift during multi-step execution.

273|45|Updated Jan 16, 2026

Search Tags:#context management #drift detection #session continuity #long-running tasks #ai safety #output fidelity

Ready to run

npx skills add https://github.com/pskoett/pskoett-ai-skills --skill context-surfing

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: context-surfing
Source: https://github.com/pskoett/pskoett-ai-skills/tree/main/skills/context-surfing
Command: npx skills add https://github.com/pskoett/pskoett-ai-skills --skill context-surfing

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill prevents AI output degradation by actively monitoring the context window's health and ensuring the AI operates at peak coherence, avoiding "wipeouts" from drift or hallucination.

Core Features & Use Cases

Context Health Monitoring: Continuously assesses the quality and coherence of the AI's context.
Drift Detection: Identifies and reacts to signals of context degradation, such as contradictions or hallucinations.
Clean Handoffs: Manages graceful exits when context quality drops, saving state for seamless resumption.
Use Case: For long-running, complex coding tasks where maintaining accurate context is critical, this skill ensures the AI doesn't silently introduce errors due to a degrading understanding of the project or task.

Quick Start

Use the context-surfing skill to monitor and maintain context quality during a multi-step coding task.

Frequently Asked Questions about context-surfing

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

FAQPage Schema

How do I prevent AI context drift during long-running coding tasks?▼

To prevent AI context drift during long-running coding tasks, you need continuous context health monitoring that detects degradation signals like contradictions or hallucinations before they corrupt output. This ensures maximum output fidelity throughout multi-step execution.

What is context window health monitoring for complex refactors?▼

Context window health monitoring for complex refactors is the continuous assessment of AI context coherence to ensure peak performance. It actively identifies degradation signals like contradictions and manages clean handoffs to maintain output fidelity.

How do I manage clean handoffs when AI context quality drops?▼

To manage clean handoffs when AI context quality drops, the system saves the current execution state during graceful exits. This allows seamless resumption of long-running tasks without losing progress or introducing silent errors.

Do I need intent-framed-agent to use context-surfing for session continuity?▼

Yes, intent-framed-agent is required for context-surfing as it provides the wave anchor. You also need plan-interview for external grounding via plan files to ensure maximum output fidelity during multi-step execution.

Why does AI output degrade in long-running tasks and how to fix it?▼

AI output degrades in long-running tasks due to context drift, where the AI's understanding silently degrades causing contradictions or hallucinations. Fix it by applying continuous context health monitoring to detect degradation and ensure output fidelity.