GitHub - hackerbara/harnessmonkey: UserScripts for Claude Code - Showing hidden tokens & thinking, stop reminders, add vibes, and more · GitHub

/" data-turbo-transient="true" />

Skip to content

Search or jump to...

Search code, repositories, users, issues, pull requests...

-->

Search

Clear

Search syntax tips

Provide feedback

--><br>We read every piece of feedback, and take your input very seriously.

Include my email address so I can be contacted

Cancel

Submit feedback

Saved searches

Use saved searches to filter your results more quickly

-->

Name

Query

To see all available qualifiers, see our documentation.

Cancel

Create saved search

Sign in

/;ref_cta:Sign up;ref_loc:header logged out"}"<br>Sign up

Appearance settings

Resetting focus

You signed in with another tab or window. Reload to refresh your session.<br>You signed out in another tab or window. Reload to refresh your session.<br>You switched accounts on another tab or window. Reload to refresh your session.

Dismiss alert

{{ message }}

hackerbara

harnessmonkey

Public

Notifications<br>You must be signed in to change notification settings

Fork

Star

main

BranchesTags

Go to file

CodeOpen more actions menu

Folders and files<br>NameNameLast commit message<br>Last commit date<br>Latest commit

History<br>1 Commit<br>1 Commit

assets/demos

assets/demos

examples

examples

packages

packages

scripts

scripts

src/harnessmonkey

src/harnessmonkey

tests

tests

.gitignore

.gitignore

LICENSE

LICENSE

README.md

README.md

pyproject.toml

pyproject.toml

uv.lock

uv.lock

View all files

Repository files navigation

HarnessMonkey - UserScripts for Claude Code

A reference example of a personal UserScript-style modification manager for Claude Code that handles applying/unapplying selected patches, command line options, and prompts to your selected claude location via a shim, patch engine, and re-packer.

Provides a Python CLI tool, (ugly) GUI, and menubar manager. Reference for Mac only currently.

Example scripts

Package<br>What it does<br>Demo

thinking-drawer<br>A footer drawer projecting the model's thinking text, raw and structured.

reminders-drawer<br>A footer drawer with live on/off toggles for seven recurring reminder/accounting attachment families. Runtime control instead of build-time suppression.

mute-reminders<br>Statically suppresses those same seven attachment families upstream. The "just make it all quiet" option. Conflicts with reminders-drawer — pick one.

hidden-context-drawer<br>A footer "Hidden Context" drawer so you can read the model-visible attachment context (reminders, timestamps, token accounting) the harness normally hides from you.

hidden-context-inline<br>Same hidden context, projected straight into the transcript as inline warning rows. Conflicts with the drawer — pick one.

heraldic-dragons<br>Two heraldic fire-breathing pixel-art dragons flanking your terminal, with animated flames. Take it to 11 sometime, you know? Needs a truecolor terminal, Ghostty etc. One art scene at a time — conflicts with capybara-onsen.

capybara-onsen<br>Capybaras chilling. It's very brave of them to do so if you think about it. Also truecolor, also one scene at a time.

fable-fallback<br>Un-hides Fable→Opus safety-classifier downgrade events: warning banner in resumed chats, marker in the /resume picker. The original reason this repo exists.

drawer-dock<br>The shared footer-drawer framework the three drawer scripts above plug into. Enable it alongside any of them. Demo shows the full dock with all three aboard.

Why these scripts?

I was tired of four things with Claude Code:

Not being able to see all the tokens the model sees

The automated reminders that fire and make Claude anxious and jumpy

Not nearly enough vibes

Needing an alias to pass my system prompt and --dangerously-skip-permissions

So these are ideas to improve my personal Claude situation, and maybe yours too. But you should think of scripts that speak to you!

Is this a good idea?

Probably not! Don't violate your Terms of Service, don't get hacked, don't crash your computer hard -- all important things to focus on in your life. Injecting arbitrary web-provided JS into an opaque agent harness with powerful permissions may interfere with these goals!

(I am neither a lawyer nor cybersecurity expert though so don't listen to me...)

How do I install?

Requires: a Mac on Apple Silicon, uv (brings its own Python), and a local Claude Code install to patch.

git clone https://github.com/hackerbara/harnessmonkey<br>cd harnessmonkey<br>uv sync<br>uv run harnessmonkey install

That's it — the monkey lands in your menubar (and comes back on login), with all the scripts loaded and switched off. From there, the three steps below.

(If the monkey doesn't appear — launch it directly: uv run harnessmonkey-gui.)

Prefer terminal-only? uv run harnessmonkey install --cli skips the menubar app; everything it does has a CLI verb (uv run harnessmonkey --help).

Changed your mind? uv run harnessmonkey uninstall takes the menubar...