Claude Code Workarounds
Unofficial community workarounds for Claude Code (empty thinking summaries; missing context-usage icon; no copy/paste of markdown).
Install
npm i -g pkgInstall command taken from the project's README or registry entry. Always review code before running it.
README excerpt
Read full README on GitHub →claude-code-workarounds
Unofficial community workarounds for Claude Code. Each entry below is an independent fix, delivered through one env-toggled launcher per platform plus standalone per-fix tools.
Not affiliated with or endorsed by Anthropic. A future Claude Code update could make any of the included workarounds obsolete. Use them at your own discretion.
Workarounds
Empty thinking summaries (Opus 4.7 / 4.8) [updated 2026-07-07]. Thinking summaries render empty in the VS Code extension and headless
-p/SDK paths, even withshowThinkingSummariesenabled. Fix via the launcher (recommended), a one-line extension patch, or a local proxy. -> details2026-07-07 - new Opus 4.8 regression, upstream and server-side. On
claude-opus-4-8only, the API now returns an empty summary even when the request correctly asks for one, so the launcher's flag is no longer sufficient on that single model. It is gated on an experiment-assignment header the CLI sends. Every other model is unaffected and the launcher remains the fix for them. There is a mitigation, and the launcher can now apply it for you: setCC_ATIS_OPTOUT=1(off by default - it has caveats). -> 2026-07-07 updateMissing context-usage icon (1M context window) [updated 2026-06-10]. The context-usage pie in the chat input is hidden until you have used more than 50% of the context window. With the 1M window that is about 500,000 tokens, so it is effectively never shown. Fix via the launcher (re-patches the webview on each launch), or a standalone patcher script. -> details
No markdown copy / export of chat [updated 2026-06-10]. The chat cannot copy a whole message or the whole conversation as Markdown, and has no transcript export. Fix via the launcher (adds copy controls, re-applied each launch), a standalone patcher, or a standalone session exporter CLI. -> details
Quick start
Two downloads, one per platform:
- Linux / macOS:
launcher/claudemax- the bash launcher, no build needed. - Windows:
claudemax.exefrom the Releases page.
Put it on your PATH, point the VS Code "Claude Code" extension's claudeCode.claudeProcessWrapper setting at its full path, and reload the window. That enables every fix above. Terminal use, per-fix toggles, and the full wiring are under The launcher and each workaround's section.
The launcher
The recommended fix for everything is one small launcher that wraps the real claude binary. It is a drop-in process wrapper carrying every fix in this repo; each fix is on by default and independently switchable with an environment variable, so the same artifact serves "I want everything" and "I want only X" without editing code and without recompiling.
- Linux / macOS: the bash script
launcher/claudemax. - Windows: the compiled
claudemax.exeon the Releases page, built fromlauncher/claudemax.win.js.
Toggles (set in the environment where Claude Code launches, then reload):
Similar plugs
CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies
Bash is all you need - A nano claude code–like 「agent harness」, built from 0 to 1
Documented system prompts from Anthropic - Claude Fable 5.1, Opus 5.5, Claude Design, Claude Code. OpenAI - ChatGPT GPT-6-Astra, Codex. Google - Gemini 3.8 Flash, 3.1 Pro,…
Wrap Antigravity, ChatGPT Codex, Claude Code, Grok Build, Muse Code, Devin as an OpenAI/Gemini/Claude/Codex compatible API service, allowing you to enjoy the free Gemini…