Skip to content

Built-in Tools Overview

ChatShell ships with 9 built-in tools that your AI uses autonomously. These tools are always available in every conversation — no MCP servers to configure, no plugins to install.

ToolDescription
Web SearchSearch the web and retrieve relevant results
Web FetchFetch and read the full content of any web page
BashExecute shell commands on your local machine
ReadRead file contents from your local filesystem
EditMake precise text replacements in files
WriteCreate or overwrite files with provided content
GrepSearch file contents using pattern matching
GlobDiscover files by name patterns
Kill ShellTerminate the current bash session and start fresh

When you send a message, the AI model decides which tools (if any) to use based on your request. It can chain multiple tool calls in sequence — for example, searching the web, fetching specific pages, and then reading a local file to compare results.

You can watch each tool call as it happens: ChatShell shows the tool name, its input, and a summary of the output in real time.

Here are three real examples of how ChatShell chains tools autonomously:

Travel planning with Web Search:

“Plan a 3-day trip to Barcelona including must-see sights and restaurant recommendations.”

ChatShell makes ~6 web searches and ~8 page fetches to gather current, specific information before producing a detailed itinerary — all autonomously.

Movie research with deep web fetching:

“Find and compare the best sci-fi movies of the last 5 years.”

18 tool calls in one query: 6 Web Searches + 12 Web Fetches to find, read, and synthesize detailed reviews and ratings across multiple sources.

Project analysis with local tools:

“Analyze this codebase and give me an architecture overview.”

20 tool calls: Glob to discover all source files, Bash to inspect build configs and scripts, and Read to examine key modules — producing a structured architecture document.

By default, all built-in tools are available. You can control which tools are active per conversation in the conversation settings panel.

For additional tools beyond the built-ins, see MCP Integration.