Skip to content
claudemods

Claude Code Custom Tool Mods: Give Claude New Tools

Claude Code custom tool mods register new tools Claude can call, written in JavaScript inside Claude Code, as a lighter alternative to running an MCP server.

>_ browser-lanes

Browser Lanes

A band above the prompt showing whether your session has a Playwright browser, who is driving it, and who waits.

  • above-prompt
  • toast
  • Reads files
  • Runs commands
  • Intercepts tool calls
  • Changes the UI
  • Adds commands/tools
  • Stores data

tested with v2.1.307

>_ ccpii

ccpii

Masks emails, names and phones with placeholders before Claude reads them, using a local privacy-filter model.

  • status
  • toast
  • transcript
  • Reads files
  • Writes files
  • Runs commands
  • Network
  • Intercepts tool calls
  • Reads your prompts
  • Changes the UI
  • Adds commands/tools
  • Stores data

tested with v2.1.310

>_ diagram-pane

diagram-pane

A side pane that pins the mermaid and ASCII diagrams Claude writes and keeps a live, editable map of the work.

  • pane
  • Writes files
  • Calls the model (uses your quota)
  • Intercepts tool calls
  • Reads your prompts
  • Changes the UI
  • Adds commands/tools
  • Stores data

tested with v2.1.307

Diff running in Claude CodeFeatured

Diff

A /diff pane for Claude Code: uncommitted git changes, file by file with their hunks, refreshed as Claude edits.

  • pane
  • Reads files
  • Runs commands
  • Intercepts tool calls
  • Reads your prompts
  • Changes the UI
  • Adds commands/tools
  • Stores data

tested with v2.1.310

>_ dream-pass

Dream Pass

Tidies Claude's memory folder: merges duplicates, rewrites stale entries, removes dead ones, with backup and undo.

  • above-prompt
  • pane
  • toast
  • Reads files
  • Writes files
  • Runs commands
  • Calls the model (uses your quota)
  • Intercepts tool calls
  • Reads your prompts
  • Changes the UI
  • Adds commands/tools
  • Stores data

tested with v2.1.307

Human in the loop running in Claude CodeFeatured

Human in the loop

Claude assigns you the tasks only you can do. They wait in a My tasks pane, counted under the prompt, until you answer.

  • pane
  • status
  • toast
  • transcript
  • Intercepts tool calls
  • Reads your prompts
  • Changes the UI
  • Adds commands/tools
  • Stores data

tested with v2.1.310

>_ memory-fork

memory-fork

Haiku checks each turn for memory-worthy facts; a tool-less fork writes them to Claude Code memory in the background.

  • above-prompt
  • pane
  • Reads files
  • Writes files
  • Runs commands
  • Calls the model (uses your quota)
  • Intercepts tool calls
  • Reads your prompts
  • Changes the UI
  • Adds commands/tools

tested with v2.1.307

>_ pii-guard

PII Guard

Replaces Taiwanese personal data with placeholders before Claude sees it and restores real values when Claude writes.

  • none
  • Reads files
  • Runs commands
  • Network
  • Intercepts tool calls
  • Reads your prompts

tested with v2.1.291

Reflect running in Claude CodeFeatured

Reflect

Finds reusable corrections in your prompts and offers them above the prompt, to save to CLAUDE.md with one press.

  • above-prompt
  • toast
  • Reads files
  • Writes files
  • Calls the model (uses your quota)
  • Reads your prompts
  • Changes the UI
  • Adds commands/tools
  • Stores data

tested with v2.1.310

>_ secret-box

secret-box

Claude Code uses an API key without seeing it: a masked box saves it to .env and tool output comes back redacted.

  • pane
  • above-prompt
  • toast
  • transcript
  • Reads files
  • Writes files
  • Runs commands
  • Intercepts tool calls
  • Reads your prompts
  • Changes the UI
  • Adds commands/tools

tested with v2.1.307

Secret Redactor running in Claude CodeFeatured

Secret Redactor

Swaps API keys, emails and public IPs for placeholders before Claude sees them, and restores them when a tool runs.

  • toast
  • Intercepts tool calls
  • Reads your prompts
  • Changes the UI

tested with v2.1.291

>_ show-your-work

Show Your Work

Claude plays screenshots and videos of its work inline in Ghostty, with before/after compare and QA checks.

  • transcript
  • pane
  • Reads files
  • Runs commands
  • Network
  • Intercepts tool calls
  • Reads your prompts
  • Changes the UI
  • Adds commands/tools
  • Stores data

tested with v2.1.296

>_ terminal-browser

terminal-browser

Opens a real web browser in a pane inside Claude Code with /browser, drawn with the kitty graphics protocol.

  • pane
  • Reads files
  • Runs commands
  • Network
  • Intercepts tool calls
  • Changes the UI
  • Adds commands/tools

tested with v2.1.310

>_ transcript-redactor

Transcript Redactor

Swaps API keys and passwords for stable placeholders before Claude reads them, and restores them for local tools.

  • toast
  • transcript
  • Reads files
  • Intercepts tool calls
  • Reads your prompts
  • Changes the UI

tested with v2.1.307

Claude Code custom tool mods give Claude new tools to call during a turn, such as a project-specific search, a lookup in your issue tracker, or a calculator for something your codebase cares about. The tool is written in JavaScript inside the mod and runs in Claude Code, so there is no separate server process to start.

How do Claude Code custom tool mods add a tool?

A mod calls $.tool.register(...) with a name and a description that Claude reads, plus the input it accepts. When Claude decides to use the tool, the call passes through the tool.call event like any built-in tool, and the mod answers it by returning { result }. $.tool.list() shows what is available and $.tool.call() lets a mod invoke a tool itself.

Two other events matter here. tool.describe fires when a tool's description is first sent to Claude; a mod can rewrite the text or mark the tool as deferred so it does not take up context until needed. agent.offer and $.agent.register do the same job for subagent types. The API reference documents the $.tool and $.agent methods.

What should you check before installing one?

A tool mod shows Adds commands/tools. The rest of its badges tell you what the tool can reach. A tool that queries a web API shows Network; one that searches files shows Reads files; one that runs a linter shows Runs commands. Because Claude decides when to call the tool, those capabilities can be used in any turn, not only when you ask, so review them as if Claude had direct access.

Check how errors are reported too. A good tool returns a clear error message as its result instead of throwing, so Claude can recover.

Tool mods versus MCP servers

MCP servers are the standard way to give Claude tools, and they work across many clients. A tool mod is narrower: it only works in Claude Code, but it needs no install step beyond the plugin, starts with the session, and can share state with the rest of the mod, for example a status line that shows how often the tool was used.

FAQ

Should I build a tool mod or an MCP server?
Use an MCP server when the tool must work with other clients or runs as a separate service. A tool mod is simpler when the tool only needs Claude Code and a few lines of JavaScript.
Do custom tools use more tokens?
Every tool description is sent to Claude, so each one adds a little to the context. Mods can mark a tool as deferred through tool.describe so its description is only loaded when needed.
Can a tool mod see my other tool calls?
Only if it also hooks tool.call. Registering a tool with $.tool.register alone does not let a mod watch other tools; the permission badges show whether it intercepts tool calls.