Hush

TL;DR — Claude bills you for every word it reads and writes, and most of those words are logs, build output, and play-by-play. hush trims them at the source, automatically. Install it, change nothing, pay less: $0.159 an average session against $0.179 without it.

What is this?

You've seen it: "Let me start by looking at the codebase." "Now I'll check the config." Four hundred lines of build output you didn't ask for, and — eventually — the one sentence you actually needed. Every word of that is billed.

hush doesn't ask Claude to "be more concise" and hope for the best. It trims the actual bulk — logs, command output, narration — at the source, before any of it hits your bill. It earns its keep in real engineering sessions, the kind that read logs and run builds, because that's where the noise lives.

Why you'd want it

• Cheaper sessions. The two biggest sources of bulk — noisy output and narration — get shrunk, so long sessions cost less.
• Easier to read. The answer sits at the top of one final message, not buried in a play-by-play.
• Nothing important is lost. Failing command output, code, diffs, and security warnings are kept whole.
• Zero setup. Install it and it's on. Tune it later only if you feel like it.

How it works

Five small habits, picked up the moment it's installed:

That's the whole list. No workflow to learn, no dial to find first — it's just how Claude behaves now.

Install

Inside Claude Code, run:

/plugin marketplace add V-Songbird/foundry
/plugin install hush@foundry

Takes effect at your next session — nothing to configure. hush works in the background.

Running razor too? Good instinct — the pair is measured in Better together below.

What you can do

hush runs itself; these commands are extras:

[!IMPORTANT] hush-compress never touches your original — it writes a copy alongside it (CLAUDE.md → CLAUDE.hush.md) for you to review and swap in yourself.

/hush:stats needs HUSH_DEBUG=1 set before the work you want measured. Without it there's nothing to report — and it says so.

/hush:pick-style is the shelf. Six styles ship with hush, same silent machinery underneath, and they differ only in what that one last message is for:

/hush:craft-style goes one further: your own voice, written to a file you own, and a verifier that checks hush's mechanics came through the rewrite — every number, every cap, every rule, one paragraph for one paragraph.

Ask for a pirate and you get a pirate — be for is, ye for you, dropped g's, the whole way through, with the paths and the error text still exact. The trick is that your voice gets written into every line of the style file, not just the line that names it: Claude answers in the register it was handed. craft-style does that part for you. Want it terser than stock's readability rules allow? Ask for maximum compression — the skill strips the readability frame Rock-style, keeps the silence and the exact-facts contract, and tells you what you traded.

Both commands ask before they swap, both take effect at your next session, and stock hush is always one command away. One honest caveat: only the built-in style is benchmarked — the presets and anything you craft are unmeasured, and the numbers on this page belong to stock.

Benchmarks

We put hush up against plain Claude Code and two rivals — caveman, which tells Claude to talk less, and an all-round "efficiency mode" plugin — on real engineering work: full agent sessions that explore, edit, and run code. Same jobs, phrased the way a developer actually types them, real cost read straight from the API.

Being brief isn't enough. Asking Claude to talk less saves under half a cent. An efficiency mode saves about one. hush saves two. Asking politely and actually doing the work are two different things.

Almost the whole bill is what Claude reads, not what it writes back. A plugin that only shortens the reply is working on two cents of an eighteen-cent session. hush trims the logs and output before they hit your bill.

It shows most in longer sessions. Drag a huge file into a multi-turn conversation and that bulk gets re-sent every turn. hush keeps a tidy summary in the chat and the full copy one click away — that outage session came in at $0.28 against $0.43. Neither rival moved it more than a cent and a half.

And you read a third less. Across the suite the replies come to 95 words against 145 — one message at the end, outcome first, instead of pieces arriving while Claude works.

And it mostly says nothing until it's done. That's the waveform at the top of this page — every session in the suite, one spike per run. hush is silent in 27 sessions out of 30. It isn't a gag order: Claude still speaks up to flag something you'd want to stop, or when it's blocked and needs you.

The full picture

Every job, every setup — the wins and the ties and losses. Cheapest per row in bold.

Every job passed its correctness check in every setup — not one cheaper-but-wrong answer.

[!NOTE] hush wins where there's noise to cut — logs, long sessions, multi-turn debugging — and roughly ties on short or low-output jobs, where a session's fixed overhead dwarfs anything a plugin can trim. On a few it costs a hair more. That's the honest shape, and it's why the average is the number to read.

How we tested: same jobs, four setups, several runs each in fresh throwaway workspaces, on Sonnet — a full multi-turn agent session every time, never a single generated reply — costs read from the API, not estimated. Numbers move a few percent between runs. Reproduce it yourself — see benchmarks/.

Better together

We ran the pair too — hush alongside razor — against the rival pair, caveman with ponytail. Ours came out cheapest on both models and was the only setup that never turned in a wrong answer; the rival pair managed to cost more than running no plugin at all. The difference is enforcement: caveman and ponytail ask — talk less, build lean — and asking works right up until the model forgets. hush and razor fire on every session, whether Claude is in the mood or not.

Under the hood

Every trim above happens locally as Claude works — read the plugin's files if you want the exact mechanics. craft-style copies those measured mechanics verbatim into a new style file in your own output-styles folder, checked by a mechanical verifier; the six shipped presets are built the same way and pass the same verifier. With your say-so pick-style swaps whichever one you chose into hush's own slot so it binds like stock, and swaps stock back on request. A plugin that takes plugins, more or less. Pairs naturally with razor: razor cuts the code, hush cuts the noise. Run both and neither notices the other — measured as a pair, they're the setup we'd pick ourselves (see Better together).

Settings

Most people never touch these. A few environment variables tune the caps or turn parts off:

License

MIT — see LICENSE.