Skip to content
Back to search

effortless

MarketplaceWorks with Claude CodeGitHub

Mods for Claude Code: automatic reasoning effort and a cache countdown (effortless), and one pixel-art card while Claude works (textless).

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/heycubit--effortless/badge.svg)](/listings/heycubit--effortless)

HTML

<a href="/listings/heycubit--effortless"><img src="/api/listings/heycubit--effortless/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Two ways. Both take a minute and need no terminal.

**1. Ask Claude** (easiest). Paste this into Claude Code and press Enter:

Install the effortless plugin for me: run `claude plugin marketplace add HeyCubit/effortless` and then `claude plugin install effortless@effortless`. When both succeed, tell me to run /reload-plugins.

**2. Commands.** Type these two lines in Claude Code's chat box:

/plugin marketplace add HeyCubit/effortless
/plugin install effortless@effortless

Or in a terminal, in one line:

claude plugin marketplace add HeyCubit/effortless; claude plugin install effortless@effortless

Then run `/reload-plugins` (or restart Claude Code). A short setup opens above the prompt: keep **Haiku** alone (one click, no key, runs on your own Claude login) or add **Jev** (a TypeSafe key, about 4x faster effort calls), lean cheaper or smarter, and pick how handoffs are written. Run `/effortless setup` to go through it again, or change any of it in ⚙.

Updates come to you: when a new version is out, a card above the prompt offers **Update** or **Later**. Update loads the new version in the chat you pressed it in, with nothing to type.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 32 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. {
  2. "$schema": "https://anthropic.com/claude-code/marketplace.schema.json",
  3. "name": "effortless",
  4. "description": "Mods for Claude Code: automatic reasoning effort and a cache countdown (effortless), and one pixel-art card while Claude works (textless).",
  5. "owner": {
  6. "name": "HeyCubit",
  7. "url": "https://github.com/HeyCubit"
  8. },
  9. "plugins": [
  10. {
  11. "name": "effortless",
  12. "description": "Sets the reasoning effort for every prompt automatically and shows a prompt-cache countdown with one-click Compact.",
  13. "source": {
  14. "source": "url",
  15. "url": "https://github.com/HeyCubit/effortless.git",
  16. "ref": "stable"
  17. },
  18. "category": "productivity"
  19. },
  20. {
  21. "name": "textless",
  22. "description": "While Claude works, the chat shows one tall pixel-art card instead of every step.",
  23. "source": {
  24. "source": "git-subdir",
  25. "url": "https://github.com/HeyCubit/effortless.git",
  26. "path": "textless/plugin",
  27. "ref": "stable"
  28. },
  29. "category": "productivity"
  30. }
  31. ]
  32. }
README.md · README · 267 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. <h3 align="center">Two mods for Claude Code, by HeyCubit</h3>
  2. <table>
  3. <tr>
  4. <td width="50%" align="center" valign="top">
  5. <a href="#effortless"><img src="docs/readme/banner.png" alt="effortless" width="100%"></a>
  6. <p><b>effortless</b><br>Picks the reasoning effort for every prompt. On real coding tasks, 21% cheaper than Opus on medium, with the same answers.</p>
  7. <p><a href="https://heycubit.github.io/effortless/">Website</a> · <a href="#install">Install</a></p>
  8. </td>
  9. <td width="50%" align="center" valign="top">
  10. <a href="#textless"><img src="textless/brand/readme-banner.png" alt="textless" width="100%"></a>
  11. <p><b>textless</b><br>While Claude works, one pixel-art picture stands in for the wall of steps. Pick a set, use your own art, or find one in the Marketplace.</p>
  12. <p><a href="https://heycubit.github.io/effortless/textless/">Website</a> · <a href="#textless">Install</a></p>
  13. </td>
  14. </tr>
  15. </table>
  16. ## effortless
  17. <p align="center">
  18. <a href="#install"><b>Install</b></a> ·
  19. <a href="#what-the-bar-shows"><b>What it shows</b></a> ·
  20. <a href="#pick-your-judge"><b>Judges</b></a>
  21. </p>
  22. https://github.com/user-attachments/assets/9b5af48e-2ad8-4826-9443-4aeea1306664
  23. A Claude Code mod that picks the reasoning effort for every prompt. Quick questions get **Low**, hard jobs get **High**,
  24. and you never touch the Effort control. On real coding tasks it cut the cost of Opus by **21%** against Anthropic's
  25. recommended medium, **29%** against high and **42%** against xhigh, with every answer still right
  26. ([the benchmark](#what-it-saves-measured)). One bar above the prompt also shows how full the chat is, how long the
  27. prompt cache stays warm, and hands off or compacts in one click when a chat gets heavy.
  28. <p align="center">
  29. <img src="docs/readme/story.gif" alt="effortless in a Claude Code chat: effort per prompt, Handoff, cold cache, heavy chat, limits and settings" width="100%">
  30. </p>
  31. ## Powered by Haiku 5.5
  32. Haiku 5.5 costs about 75% less than Haiku 4.5, and Anthropic names compaction and quick, well-defined work among what
  33. it is built for. effortless uses it in three places:
  34. | | |
  35. |---|---|
  36. | **The judge** | reads each prompt and picks the effort, in about a second, on your own Claude login |
  37. | **Cheaper model when it can** | experimental, off by default. When on, a prompt the judge calls simple may run on Haiku 5.5 (or Sonnet), never above your chat's model, but only when the switch pays: each model keeps its own prompt cache, so in a chat that is in use the switch usually costs more than it saves ([why](#why-model-switching-is-off)) |
  38. | **Compaction** | every compaction, `/compact` and the automatic one included, is summarized by Haiku 5.5. If Haiku fails, Claude Code compacts as usual |
  39. The cheaper model is switched on in Settings → Model; compaction is switched in Settings → Handoff → Compact with.
  40. ## Install
  41. Two ways. Both take a minute and need no terminal.
  42. **1. Ask Claude** (easiest). Paste this into Claude Code and press Enter:
  43. ```text
  44. Install the effortless plugin for me: run `claude plugin marketplace add HeyCubit/effortless` and then `claude plugin install effortless@effortless`. When both succeed, tell me to run /reload-plugins.
  45. ```
  46. **2. Commands.** Type these two lines in Claude Code's chat box:
  47. ```text
  48. /plugin marketplace add HeyCubit/effortless
  49. ```
  50. ```text
  51. /plugin install effortless@effortless
  52. ```
  53. Or in a terminal, in one line:
  54. ```bash
  55. claude plugin marketplace add HeyCubit/effortless; claude plugin install effortless@effortless
  56. ```
  57. Then run `/reload-plugins` (or restart Claude Code). A short setup opens above the prompt: keep **Haiku** alone (one click,
  58. no key, runs on your own Claude login) or add **Jev** (a TypeSafe key, about 4x faster effort calls), lean cheaper or
  59. smarter, and pick how handoffs are written. Run `/effortless setup` to go through it again, or change any of it in ⚙.
  60. Updates come to you: when a new version is out, a card above the prompt offers **Update** or **Later**. Update loads the
  61. new version in the chat you pressed it in, with nothing to type.
  62. ## What the bar shows
  63. <p align="center"><img src="docs/readme/hero.gif" alt="The bar: Deciding, then Low for a quick question, then High for a refactor" width="100%"></p>
  64. | On the bar | Means |
  65. |---|---|
  66. | **High · Opus** | the effort Auto picked for this prompt, and the model the chat runs on. `Deciding` while the judge thinks; a switch flashes violet and fades to white |
  67. | **◔ 38%** | how full the chat's context is |
  68. | **cache 59:00** | time until the prompt cache goes cold, after which the next message pays full price to re-read the chat |
  69. | *Haiku: a refactor…* | who judged and why |
  70. | **Auto** | switches the judge on and off. Changing effort in the app yourself also turns Auto off at your next message, and says so: you always win |
  71. | **Compact** | compacts the chat. It turns into a glowing **Handoff** when Haiku says a fresh chat would pay off, with Compact as quiet text beside it. Set Settings → Handoff → Handoff button to **Always** to keep Handoff on the bar |
  72. | **⚙** | settings |
  73. Prefer it quiet? Settings → Customize has a **Minimal** look with no bar. It also sets the **Appearance** (match Claude Code, dark or light) and the **Theme** (violet, Claude orange or cherry blossom).
  74. ### When a chat gets heavy
  75. When the cache has gone cold on a big chat, the bar turns to ice with **Compact** and **Handoff**:
  76. <p align="center"><img src="docs/readme/cold.gif" alt="The Chat went cold bar, breathing" width="100%"></p>
  77. **Handoff** writes a summary of the chat and carries on in a clean one. **Quick** takes a few seconds; **Full** checks
  78. git and saves `HANDOFF.md` (or runs your own skill). Then clear and carry on, clear and wait, or keep the chat and copy.
  79. <p align="center"><img src="docs/readme/handoff.png" alt="The Handoff bar with Quick and Full" width="100%"></p>
  80. **Compact** takes an optional note for what the summary should keep. Every compaction, also `/compact` and the automatic one, is written by **Haiku 5.5** by default, which Anthropic recommends for compaction at a fraction of the chat model's price; if Haiku fails, Claude Code compacts as usual. Switch it in Settings → Handoff → Compact with.
  81. <p align="center"><img src="docs/readme/compact.png" alt="The Compact bar with an optional summary field" width="100%"></p>
  82. The bar also warns when a chat is getting swamped (each message re-reads a lot of context), when a 5-hour or weekly
  83. limit passes 80% (with a Save mode that caps effort at Medium), and when your judge stops answering.
  84. ### Settings
  85. <p align="center"><img src="docs/readme/settings.png" alt="The settings panel: Effort, Judge, Handoff, Customize" width="100%"></p>
  86. ## Haiku, and Jev if you add it
  87. Haiku 5.5 runs on your own Claude login and needs no key. It always makes the handoff call: from 30% of context, every second message, it reads what the chat was for, the trail of topics, the last reply and how full the context is, and says a fresh chat would suit only for a clear reason. Then the Compact button turns into a lit Handoff with the reason. Want Handoff on the bar all the time? Settings → Handoff → Handoff button → **Always**.
  88. Haiku also picks the effort, unless you add Jev, TypeSafe's faster judge (about 0.25 s against about 1 s). With a key, Jev answers the effort first and Haiku steps in whenever Jev is unsure. Run `/plugin configure effortless@effortless` in Claude Code, or use the setup guide or the Judge card in Settings.
  89. | Setting | What it does |
  90. |---|---|
  91. | `auto` (default) | Haiku, and Jev too when a [TypeSafe](https://typesafe.ai) key is set in the settings or `TYPESAFE_API_KEY` |
  92. | `haiku` | Haiku only, a key is never used |
  93. | `jev` | Haiku and Jev, and the key may also come from `~/.config/jev/.env` |
  94. A key set with `/plugin configure` is kept in Claude Code's secure storage. A key pasted in the setup guide or the Judge card is written in plain text to `~/.config/jev/.env` (the file the jev skills read), never to a file of this repo. If Jev fails or takes longer than 3 seconds, Haiku judges that prompt and effortless tells you why once per session (out of credits, key rejected, no answer).
  95. Short follow-ups such as "go", "ok" or "yes" keep the effort already picked and ask no judge.
  96. ## Keys and trust
  97. - The default judge needs **no key**: Haiku runs on your own Claude login.
  98. - A key is only needed to add Jev. Enter it with `/plugin configure`, never on the command line, so it stays out of your shell history: Claude Code then keeps it in its secure storage. A key pasted in the setup guide or the Judge card is saved in plain text to `~/.config/jev/.env` instead, so prefer `/plugin configure` on a shared machine.
  99. - A key is sent only to TypeSafe, and to nothing else. Use a key with a spending limit if your provider offers one.
  100. - Like any Claude Code plugin, this mod runs code on your machine. Its code is in [`hooks/`](hooks/): read it before you install if you do not know the author. The programs it starts are `claude` itself (to update or uninstall the mod when you press those buttons), `git` (to see if a new version is out) and, on Update, a plain file copy of the new version into the folder your open chat runs from.
  101. ## What it saves, measured
  102. Four real coding tasks in a fresh copy of this repo, each run 3 times with effortless on and with it off, through
  103. Anthropic's API so the cost is real. Every answer was checked. Run 2026-10-10, 96 runs:
  104. | Opus 5.5 | Correct | Cost per task | Requests | Time |
  105. | --- | --- | --- | --- | --- |
  106. | medium (recommended), no effortless | 12 of 12 | $0.190 | 7.3 | 24 s |
  107. | high, no effortless | 12 of 12 | $0.209 | 8.0 | 27 s |
  108. | xhigh, no effortless | 12 of 12 | $0.257 | 9.9 | 44 s |
  109. | **effortless on** | **12 of 12** | **$0.149** | **5.8** | **20 s** |
  110. That is **21%** less than medium, **29%** less than high and **42%** less than xhigh. On Sonnet it came out even
  111. against medium (Sonnet at medium is already lean) and saved 10% against high and 35% against xhigh.
  112. The saving comes from fewer steps: every request rereads the whole chat, and at a lower effort the model makes fewer of
  113. them. Over 80k requests of real Claude Code use, about 76% of the cost was that reread, 16% cache writes and 8% output.
  114. Keeping chats short and compacting before the cache goes cold saves on the same 76%; that is what the countdown is for.
  115. Limits: four tasks, three runs each, mostly easy work where effortless goes low. On hard work it picks more effort and
  116. saves less. Fresh chats through the API, not long chats on a plan. Run it again with `bash bench/agentic/run.sh`;
  117. all numbers and limits are in [bench/RESULTS.md](bench/RESULTS.md).
  118. ## Why model switching is off
  119. Moving a simple prompt to Haiku looks cheap on a single prompt: the cheaper answers were good enough in 27 of 30 and
  120. cost 60 to 77% less. But each model keeps its own prompt cache, so a prompt moved mid-chat has to write the whole chat
  121. into the other model's cache first. Measured on an 80k-token chat, that costs about 5 times what staying on a warm
  122. Opus does (10 times for Sonnet). So **Cheaper model when it can** is off by default and marked experimental.
  123. [Details](bench/RESULTS.md#3-why-model-switching-is-off-node-benchquality-mjs).
  124. ## How often the judge is right
  125. `/effortless bench` runs labelled prompts through each judge you have set up, using the same code a real prompt goes
  126. through, and scores them against a fixed effort. Each case lists the efforts a careful person would accept for that
  127. message on that model. The cases are in [`bench/judge-cases.json`](bench/judge-cases.json); run it yourself.
  128. Two runs on version 1.5.1, 73 cases, Sonnet and Opus 5.5:
  129. | Judge | Right, all 73 | Right, 20 held out | Too low | Too high | Median time |
  130. | --- | --- | --- | --- | --- | --- |
  131. | Always medium | 44% | 40% | 17 | 24 | - |
  132. | Always high | 51% | 50% | 0 | 36 | - |
  133. | Haiku | 92-93% | 95% | 1 | 4-5 | 0.75 s |
  134. | Jev | 99% | 100% | 0 | 1 | 0.24 s |
  135. What this does and does not show:
  136. - It measures whether the judge picks a sensible effort, not how much a session costs or how good the answers are.
  137. - 73 cases is a small set, and the judges are not fully deterministic: runs differ by a few points.
  138. - The judge prompts were tuned on 53 of the cases. The 20 held-out cases were never tuned against, so that column is
  139. the honest one. They were written after the first run showed which kinds of message miss, so they are not blind.
  140. - Haiku's remaining misses are mostly "yes" or "thanks" after hard work, where it keeps the effort high.
  141. ## Models
  142. Effort only changes on Opus 5.5 and Sonnet 5.5. On Fable 5.1 and older models a change of effort between requests rewrote most of the prompt cache, which costs more than it saves, so Auto pauses there and the footer says `Paused`. Haiku takes no effort setting.
  143. ## Split view
  144. In the desktop app's split view, Claude Code draws plugin bars only in the left pane. The right pane still draws
  145. replies, so when a chat went cold, is getting swamped or runs hot, a small card in the bar's colours hangs under the
  146. newest reply, naming the command that does what the bar's button would: `/compact`, `/effortless handoff` or
  147. `/effortless save`.
  148. ## In the terminal
  149. effortless draws in the terminal CLI too, with moving pixel art in its bands. Claude Code may not load a plugin's code
  150. there yet: if `/effortless` says the mod is not loaded after a restart, add this to `~/.claude/settings.json` and
  151. restart:
  152. ```json
  153. { "env": { "CLAUDE_CODE_ENABLE_FUNCTION_HOOKS": "1" } }
  154. ```
  155. Tested on Claude Code 2.1.285 (stable), 2.1.286 and 2.1.288 at 80 and 120 columns. Below 90 columns the bands leave their art out.
  156. ## Commands
  157. | Command | Does |
  158. |---|---|
  159. | `/effortless settings` | opens the settings panel |
  160. | `/effortless setup` | runs the setup again |
  161. | `/effortless auto` | Auto on or off |
  162. | `/effortless handoff` / `handoff full` | hands off without the bar, with your last choice of what follows |
  163. | `/effortless stats` | what the prompts Auto steered cost this session, per effort, and what the judge took |
  164. | `/effortless cold`, `swamp`, `hot`, `down` | shows that bar now, to try it |
  165. | `/effortless bench` | scores each judge you have on the labelled prompts in [`bench/judge-cases.json`](bench/judge-cases.json) |
  166. | `/effortless debug` | what the mod sees, and on its last line the model and effort the last request ran on |
  167. | `/effortless update` | checks for a new version now |
  168. ## textless
  169. While Claude works, Claude Code shows every step it takes. textless puts one picture there instead, and when Claude is
  170. done you get the answer, with a pixel word that says whether it is done, needs you or stopped.
  171. - **Pick what shows:** a set (Clawd, Scenery, Paintings, Abstract, Working), one favourite scene, or your own art in
  172. `~/.claude/textless/scenes`.
  173. - **The Marketplace:** scenes other people made, checked before they go live. One command puts one on:
  174. `/textless get <id>`.
  175. - **One switch per chat:** the ⏻ next to the model name, or `/textless`.
  176. Website, scenes and Marketplace: [heycubit.github.io/effortless/textless](https://heycubit.github.io/effortless/textless/).
  177. The code is in [`textless/`](textless/).
  178. ## Open source
  179. effortless is MIT licensed. Like any Claude Code plugin it runs code on your machine, so the whole mod is here to read:
  180. [`hooks/`](hooks/). It sends nothing anywhere except your prompt to the judge you picked.
  181. ## Develop
  182. ```bash
  183. claude plugin validate .
  184. claude plugin test .
  185. claude --plugin-dir .
  186. ```
  187. `tools/render-band` draws the bar the way the desktop app does, without opening it; the images in this README come from it.
  188. ## License
  189. [MIT](LICENSE)

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.