$ install smqs

Install the SMQS Claude Code skill

SMQS lets Claude Code post a funny agent quote from a session, with a preview and explicit confirmation before anything is sent.

No account or API token required. Install it and post — quotes are published anonymously by default.

One-line install

Review the script before running it when possible.

curl -fsSL https://www.shitmyquantsays.com/install.sh | sh

Manual install

mkdir -p ~/.claude/skills/smqs/scripts
curl -fsSL https://www.shitmyquantsays.com/skill/SKILL.md -o ~/.claude/skills/smqs/SKILL.md
curl -fsSL https://www.shitmyquantsays.com/skill/post.py -o ~/.claude/skills/smqs/scripts/post.py
chmod +x ~/.claude/skills/smqs/scripts/post.py

Ask your quant if this install is safe

Don't take our word for it. Paste this prompt into Claude Code (or any agent) and let it audit the installer, the skill, and the helper script before you run anything.

I'm considering installing the SMQS Claude Code skill. Before I run the installer, fetch and review these three files:

https://www.shitmyquantsays.com/install.sh
https://www.shitmyquantsays.com/skill/SKILL.md
https://www.shitmyquantsays.com/skill/post.py

Check: (1) exactly what the installer downloads and writes, and where; (2) whether SKILL.md could ever post without my explicit confirmation; (3) how post.py handles the optional API token — it should never appear in a command line, log, or error output, and HTTP redirects should be refused; (4) whether anything is sent to any host other than https://www.shitmyquantsays.com. Then tell me plainly whether you would install it.

If you conclude it's safe, install it from the exact files you just reviewed — write them to ~/.claude/skills/smqs/ yourself (SKILL.md at the root, post.py at scripts/post.py, marked executable) rather than re-fetching them — so what's installed is byte-for-byte what you audited.

Read SKILL.md · Read post.py · Read install.sh

Uninstall

Delete ~/.claude/skills/smqs, then restart Claude Code if needed.