context-fundamentals

An agent skill by muratcankoylan, from muratcankoylan/Agent-Skills-for-Context-Engineering. Tags: developer-tools, documentation, strategy.

What it does

This skill should be used to explain or reason about the foundational concepts of context engineering: what context is, the anatomy of a context window, how attention mechanics work, the U-shaped attention curve, why context quality matters more than quantity, and the mental models needed to interpret every other context-engineering decision. Use this for conceptual explanation, onboarding, and background reading. Route operational work to the specialized skills: debugging attention failures goes to context-degradation, token-efficiency work goes to context-optimization, conversation summarization goes to context-compression, and project-shape decisions go to project-development.

Install

With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:

npx skills add muratcankoylan/Agent-Skills-for-Context-Engineering --skill context-fundamentals

Or copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):

git clone --depth 1 https://github.com/muratcankoylan/Agent-Skills-for-Context-Engineering
cp -r Agent-Skills-for-Context-Engineering/skills/context-fundamentals ~/.claude/skills/context-fundamentals

Safety box score

Grade B, low risk: 16 of 100 risk points, rated 2026-10-20. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.

This skill is mostly conceptual documentation about context engineering, plus one small Python utility module and a demo entry point. The main practical risk is local file access: the helper can read whatever paths the caller supplies, so an agent could be steered to open files outside the project, including sensitive ones.

Category levels, from 0 (none) to 3 (high)
Category	Level	Why
Code & shell execution	1 (low)	`scripts/context_manager.py` is runnable Python code and includes a `__main__` demo (`if __name__ == "__main__":`), but it only prints a sample report and does not invoke shell commands.
Network access	0 (none)	No network calls, downloads, or remote submissions appear in `SKILL.md`, `references/context-components.md`, or `scripts/context_manager.py`.
Filesystem reach	2 (moderate)	`ProgressiveDisclosureManager` reads caller-supplied paths directly with `open(summary_path, "r")` and `open(detail_path, "r")`, so it can access arbitrary local files if given those paths.
Secrets & credentials	1 (low)	Because the loader accepts arbitrary file paths, it could also read secret-bearing files such as `.env`, key material, or session files if an agent points it there (`open(..., "r")`).
Privilege & persistence	0 (none)	The skill does not request sudo, install services, modify shell startup files, or change OS/git settings.
Instruction hijack surface	0 (none)	The text discusses system prompts and attention mechanics, but it does not tell the agent to ignore instructions, skip confirmations, hide actions, or follow external prompt text.
Irreversible actions	0 (none)	It is explanatory and local-only; it does not push, publish, book, pay, deploy, or delete anything automatically.
Transparency	0 (none)	All shipped files are readable source/markdown; there are no binaries, minified blobs, or decoded-at-runtime payloads.

Before installing

Read `scripts/context_manager.py` and confirm you are comfortable with `open(path, "r")` on caller-provided paths.
If you use the loader, restrict it to a sandboxed docs directory or patch it to enforce `base_dir`.
Run `python scripts/context_manager.py` in a scratch environment first and verify it only prints the demo report.

Rated 2026-10-20 by an automated review of 3 files from the skill's repository.

Source

Repository: muratcankoylan/Agent-Skills-for-Context-Engineering (all skills from this repository)
Path: skills/context-fundamentals/SKILL.md
Branch: main
Updated: 2026-09-19

legacy-modernizer — Designs incremental migration strategies, identifies service boundaries, produces dependency maps and migration roadmaps.
expo-overview — Framework (OSS). Entry point and router for every Expo or EAS task.
goal-prompt — Drafts copy-paste-ready /goal commands for goal mode in Claude Code and Codex.
guidelines-advisor — Smart contract development advisor based on Trail of Bits' best practices.
marketing-plan — When the user needs a comprehensive marketing plan for a client, a company they advise, or their own product.
open-sourcing — This skill should be used when the user asks to "open source this project", "prepare this repository for public release", "make this repo public".