How to inject context in Claude Code using Hooks

Описание: Claude Code hooks are a hot topic in the AI-assisted development space, and for a good reason. Most developers rely on prompts and CLAUDE.md to guide Claude's behavior. The problem is that prompts are suggestions — Claude can forget them, skip them, or deprioritize them mid-session. Hooks guarantee behavior at the infrastructure level.

In simple terms, a hook is a shell script that fires automatically at a specific lifecycle event - before a tool runs, after an edit, or when a session starts. In this post (and the accompanying video), I want to focus on one powerful pattern: injecting context into Claude's reasoning for a specific command.
Suppose you have API design guidelines your team follows - consistent response shapes, zod validation, proper HTTP status codes. Without hooks, you paste these rules manually every session and hope Claude remembers. With a PreToolUse hook, you write a small script that detects when Claude is about to edit any file under src/api/, reads your guidelines from a markdown file, and injects them as additionalContext before the edit executes. Claude receives your rules at precisely the right moment, every single time.

The setup involves three pieces: a guidelines markdown file in .claude/hooks/, a bash script that reads the tool input via jq, checks the file path, and returns a JSON object with your context, and a registration entry in .claude/settings.json under the PreToolUse event with an Edit|Write matcher. Please refer to the video for the complete walkthrough.