Context Engineering

Audits and reorganises what an AI coding agent carries in its context window: instruction files, path-scoped rules, skills, subagents, hooks and the running conversation. Use when someone says "the agent ignores our CLAUDE.md", "set this repo up for Claude Code / Codex / Cursor", "my AGENTS.md is too long", "quality drops in long sessions", "it keeps inventing APIs", "what should go in a skill versus the rules file", or asks how to hand work to a subagent. Grounded in the documented loading behaviour of Claude Code, with the equivalents for Codex, Gemini CLI and Cursor.

213 stars 34 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add TerminalSkills/skills --skill context-engineering
Manual install
git clone --depth 1 https://github.com/TerminalSkills/skills.git /tmp/skills
mkdir -p ~/.claude/skills
cp -r /tmp/skills/skills/context-engineering ~/.claude/skills/context-engineering

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Context Engineering

## Overview

Context engineering is deciding what an AI coding agent has in its context window at each step: standing instructions, skill and tool listings, files it has read, command output, and the conversation so far. The window is finite and everything in it competes for the model's attention, so the work is curation: load what the current step needs, keep the rest reachable but out.

This skill does three jobs. It audits what a project loads today, moves each piece of knowledge to the mechanism that loads it at the right moment, and sets the habits that keep a long session usable. The facts below come from the tools' own documentation, not from folklore: where a number appears, it is one the vendor publishes.

## Instructions

### 1. See what is loaded today

Inventory the instruction files from the shell:

```bash
wc -l CLAUDE.md CLAUDE.local.md AGENTS.md GEMINI.md .claude/CLAUDE.md .claude/rules/*.md 2>/dev/null
find . \( -name CLAUDE.md -o -name AGENTS.md -o -name GEMINI.md \) -not -path "*/node_modules/*"
ls .claude/skills .claude/agents .cursor/rules 2>/dev/null
```

Then ask the user to run the tool's own report, because only the tool knows what it really loaded:

| Tool | Check | Shows |
|---|---|---|
| Claude Code | `/context` | token use by category, including the list under **Memory files** |
| Claude Code | `/memory` | every instruction file location and the auto memory folder |
Read the full skill on GitHub

Details

Skill name
context-engineering
Path
skills/context-engineering/SKILL.md
License
Apache-2.0
Repo updated
8 November 2026

FAQ

What does the Context Engineering skill do?

Audits and reorganises what an AI coding agent carries in its context window: instruction files, path-scoped rules, skills, subagents, hooks and the running conversation. Use when someone says "the agent ignores our CLAUDE.md", "set this repo up for Claude Code / Codex / Cursor", "my AGENTS.md is too long", "quality drops in long sessions", "it keeps inventing APIs", "what should go in a skill versus the rules file", or asks how to hand work to a subagent. Grounded in the documented loading behaviour of Claude Code, with the equivalents for Codex, Gemini CLI and Cursor.

How do I install the Context Engineering skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add TerminalSkills/skills --skill context-engineering. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Context Engineering popular?

It lives in TerminalSkills/skills, which has 213 GitHub stars and 34 forks.

Related skills

3dsmax Rendering

213

TerminalSkills/skills

Scripts and configures production rendering in Autodesk 3ds Max with the V-Ray and Corona renderers: output size and files, render elements, denoising, light mix, batch and command-line rendering, and network rendering. Use when a user asks to set up a production render, render several cameras in one batch, render from the command line or on a render farm, add render passes for compositing, or cut render time for archviz and product shots.

Other34

3dsmax Scripting

213

TerminalSkills/skills

Covers scripting Autodesk 3ds Max, the 3D modeling and rendering application, with MAXScript and Python (pymxs): scene manipulation, object creation, material assignment, camera and light setup, batch operations, and file I/O. Use when tasks involve automating repetitive 3ds Max workflows, batch processing scenes, running scripts headless with 3dsmaxbatch, creating custom tools, or scripting scene setup for archviz, product visualization, or VFX.

Productivity & Notes34

3proxy

213

TerminalSkills/skills

3proxy is a small open-source proxy server that runs HTTP/HTTPS, SOCKS4/5, SNI and TCP/UDP port-mapping proxies from one config file. Use when a user asks to set up an HTTP or SOCKS5 proxy, add proxy users and passwords, write 3proxy access rules, chain or rotate upstream (parent) proxies, limit bandwidth, connections or monthly traffic per user, run 3proxy in Docker, or fix a 3proxy.cfg that will not start.

DevOps & Cloud34

Context7

213

TerminalSkills/skills

Context7 fetches current, version-specific documentation and code examples for software libraries and hands them to an AI coding agent, so generated code matches the library version in use instead of stale training data. Use when someone says "use context7", "get the latest docs for this library", "look up the current API", "the agent keeps using a deprecated method", "set up Context7 in Claude Code or Cursor", or asks how to call a framework whose API changed recently. Covers the ctx7 CLI, the MCP server and its two tools, library IDs and version pinning, API keys and rate limits, and the RES

AI, LLMs & Agents34

Conf Pipeline Orchestration

200

lyndonkl/claude

The multi-agent communication and orchestration discipline for a staged pipeline whose stages each carry calibrated uncertainty. Agents talk through files with defined schemas and status fields (never free text); the orchestrator passes input paths plus an explicit output path plus a return contract, verifies each artifact's confidence before advancing a stage gate, holds and integrates worker outputs rather than passing them through, freezes its protected control logic against a checksum, maintains a diversity floor, hardens the fragile handoffs not the resilient hubs, and names the substitut

AI, LLMs & Agents25

Continue Dev

213

TerminalSkills/skills

Continue is an open-source AI code assistant for VS Code, JetBrains and the terminal (the cn CLI) that works with any LLM: Claude, GPT, Gemini, Ollama and other local models. Use when a user asks to configure Continue with config.yaml, add chat or tab autocomplete models, write rules or custom prompts, connect MCP servers, run cn headless in scripts, or migrate from config.json. Continue was acquired by Cursor in June 2026: the release from that month is final and the Hub is offline, so the skill covers local, self-configured use.

AI, LLMs & Agents34

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.