Ask HN: Anyone have solution to Opus verbosity in Claude Code? | Hacker News Reader

Ask HN: Anyone have solution to Opus verbosity in Claude Code?

5 pointsfiremelt13 comments

I find opus really talk much but speak nothing and usually I reply those verbosity with tldr pls and the text become readable. but I wonder if that would hiding some important info

anyone have tips or better solution for this?

View on HN

The 3 lines below help a bit. It still gets lost after the context fills up, but it mostly solves the verbosity issue,

- When responding to the user, respond in ASD-STE100 English. You can extend the dictionary if you need, but respect ASD-STE100 rules as much as you can.

- When writing comments, technical documentation and internal documents (roadmaps, plans, design references) use ASD-STE100 English. You can extend the dictionary if you need, but respect ASD-STE100 rules as much as you can.

- When writing plans: no timeline estimates, no complex code. unless specifically requested, no deployment sequence. use pseudo-code only for non-trivial operations where clarifying complexity is crucial

I've resorted to adding instructions like, "do not insert any comments unless they will be exceptionally useful in clarifying blocks of code for future developers."

Adding it to AGENTS.md seems to do nothing.

Same. Claude consistently ignores my instructions related to comments, and continues to be extremely verbose, often irrelevant, or too specific in terms of the current task. When called out on it, it says sorry.

I've resorted to having a final pass where comments are evaluated separately. This is much more successful than trying to get it to write good comments while writing the code.

I like:

"Explain in 10 sentences or fewer."

"Give me the gist of what you just said."

"ELI5"

This is what you want: https://github.com/juliusbrussee/caveman

You can add instructions in the CLAUDE.md/AGENTS.md file to modify output behavior like:

    # Response Style

    Keep routine answers concise. Lead with the result, include material caveats and the next action, and omit introductions, repetition, generic reassurance, and optional background. Expand only when the task requires explanation.

While writing a document I got good mileage out of doing something like this for every paragraph:

    Let's analyze this paragraph:
     - List the main points.
     - Can any points be removed? (Do any points need to be added?)
     - Reword each point. Be as succinct and clear as possible.

I'm not sure how many extra tokens it would consume to do this for every output. But it may be better than just asking for "TLDR pls"

(I've actually never used Claude Code. I've never had problems with verbosity in harnesses like OpenCode or T3-code. And recently I've been using Codex, not Claude models.)

no, you can't. Opus 5 decides by itself that it is better to ignore those instructions.

I've noticed Opus 5 repeatedly ignored my Claude.md in several ways. Particularly when it comes to verbosity of comments in code.

"reply in fewest words. to the point" usually helps

i currently use 4.8 on xhigh and it's tolerable

ignore opus 5. 4.6 is legit, opus 5 is a random brat

the solution is to migrate to another model, Opus 5 is terrible.

ELI18 TLDR

Reply on news.ycombinator.com