Read

The Read tool reads the contents of a local file and makes it available to the AI. It supports text files, images, and PDFs.

What It Does

• Reads the full content of a file at a given path
• Optionally reads a specific line range for large files
• Returns file content as text for the AI to analyze

Supported Formats

Read works with a wide range of file types:

• Text files: Source code (.rs, .ts, .py, .go, .js, .jsx, .tsx, etc.), configuration files (.toml, .yaml, .json, .env), documentation (.md, .txt, .rst), data files (.csv, .xml, .html)
• Images: PNG, JPEG, GIF, WebP, BMP, ICO, SVG — returns metadata (dimensions, size) and base64-encoded data
• PDFs: Extracts and returns the text content

When It’s Used

The AI uses Read when:

• You attach a file to the conversation
• It needs to inspect a specific file found via Glob
• You ask it to analyze, summarize, or explain a file’s contents

Example

“Read my Cargo.toml and tell me which dependencies might have security concerns.”

The AI will read the file at the specified path and analyze the dependency list.

Tips

• You can drag and drop files directly into the ChatShell input to attach them — the AI will read them automatically
• For large files, the AI may read specific sections using line ranges