context-engineering

Configures rules files, context hierarchy, and budget management for AI coding agents.

Updated Sep 8, 2026
One-click install
npx skills add https://github.com/sasidhar4444/ai-receptionist --skill context-engineering-sasidhar4444
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: context-engineering
Source: https://github.com/sasidhar4444/ai-receptionist/tree/main/agent-skills/skills/context-engineering
Command: npx skills add https://github.com/sasidhar4444/ai-receptionist --skill context-engineering-sasidhar4444

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? AI coding agents hallucinate APIs, ignore project conventions, and degrade in quality during long sessions because they receive too little, too much, or stale context. This Skill provides a systematic method for curating what the agent sees at each stage of a session. ## Core Features & Use Cases - Context Hierarchy Setup: Structures context into five levels from persistent rules files (CLAUDE.md, .cursorrules, AGENTS.md) down to transient conversation history, with templates for each. - Context Budget Management: Defines a 75% capacity trimming threshold, cut-first/protect lists, and compression-before-deletion strategies to prevent abrupt quality drops in long sessions. - Confusion Management: Provides explicit patterns for surfacing spec conflicts, missing requirements, and inline planning instead of silently guessing. - Use Case: When starting a new project for AI-assisted development, use this Skill to write a rules file covering tech stack, commands, conventions, and boundaries so the agent follows project patterns from the first prompt. ## Quick Start Ask the agent to create a CLAUDE.md rules file for your project covering tech stack, build commands, code conventions, and boundaries using the context engineering guidelines.

Frequently Asked Questions about context-engineering

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

FAQPage Schema
How do I set up a rules file for AI coding agents?▼

Create a CLAUDE.md file at the project root covering tech stack, build and test commands, code conventions, and boundaries. Equivalent files include .cursorrules for Cursor, .windsurfrules for Windsurf, and AGENTS.md for OpenAI Codex.

How to prevent AI agent quality from degrading in long sessions?▼

Start trimming context at 75% window capacity rather than waiting until it is full. Cut past failed attempts and verbose tool output first, protect the current task definition and active error, and compress long explorations into one-sentence summaries before deleting.

What is the difference between CLAUDE.md, .cursorrules, and AGENTS.md?▼

They are equivalent persistent rules files for different tools: CLAUDE.md for Claude Code, .cursorrules or .cursor/rules for Cursor, .windsurfrules for Windsurf, and AGENTS.md for OpenAI Codex. All define project-wide conventions loaded every session.

Why does my AI agent invent APIs that do not exist?▼

Hallucinated APIs indicate context starvation: the agent lacks your rules file and relevant source files. Load the files being modified, related tests, and one existing example of the pattern before each task so the agent references real project code.

How much context should I give an AI coding agent per task?▼

Aim for under 2,000 lines of focused, task-relevant context. Loading more than 5,000 lines of non-specific context causes the agent to lose focus; include only the files to modify, related tests, and one pattern example.

When should I start a fresh session instead of continuing?▼

Start fresh when switching between major features or when accumulated context references outdated patterns and deleted code. Before switching, summarize progress so the new session begins with a clean, accurate state.