Universal Claude.md
Ranked 17 of 23 agent skills by recommenders
Instruction set that measurably cuts Claude's output verbosity, though critics say it may hurt quality on complex tasks and is not benchmarked for accuracy.
Instruction set that measurably cuts Claude's output verbosity, though critics say it may hurt quality on complex tasks and is not benchmarked for accuracy.
25 people gave an opinion
- recommend it
- 10
- criticize it
- 15
- approval share
- 40%
Recommended in 1 of 13 months
- November 2025: not recommended
- December 2025: not recommended
- January 2026: not recommended
- February 2026: not recommended
- March 2026: recommended
- April 2026: not recommended
- May 2026: not recommended
- June 2026: not recommended
- July 2026: not recommended
- August 2026: not recommended
- September 2026: not recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 10 people recommend itNeeds 10 or more, met.
- 0.6 recommenders for every criticNeeds 2 to 1 or better, not met.
- Recommended in 1 of 13 monthsNeeds 4 or more, not met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Useful instructions that reduce verbosity
“If you're goal is quality outputs, then it is likely too extreme, but there are otherwise useful instructions in this repo to (quantifiably) reduce verbosity.”
Comment from 31 Mar 2026 in “Universal Claude.md – cut Claude output tokens by 63%” “While I love this set of prompts, I've not seen my clause opus 4.6 give such verbose responses when using Claude code.”
Comment from 31 Mar 2026 in “Universal Claude.md – cut Claude output tokens by 63%”
What people warn about
Each mark is one comment
Unclear quality impact without benchmarks
“without comparing before/after performance on some standard LLM benchmark to see if the instructions hurt intelligence”
Comment from 31 Mar 2026 in “Universal Claude.md – cut Claude output tokens by 63%” “The benchmark is totally useless. It measures single prompts, and only compares output tokens with no regard for accuracy.”
Comment from 31 Mar 2026 in “Universal Claude.md – cut Claude output tokens by 63%” “the benchmarks here are heavily biased towards single-shot explanatory tasks”
Comment from 31 Mar 2026 in “Universal Claude.md – cut Claude output tokens by 63%” May hurt quality on complex tasks
“forcing short responses can hurt coherence in longer agentic sessions”
Comment from 31 Mar 2026 in “Universal Claude.md – cut Claude output tokens by 63%” “For complex tasks this is not a useful prompt.”
Comment from 31 Mar 2026 in “Universal Claude.md – cut Claude output tokens by 63%” “this would, generally, make the model worse”
Comment from 31 Mar 2026 in “Universal Claude.md – cut Claude output tokens by 63%” Could raise total token spending
“Considering the 1:5 (input:output) token cost ratio, this could increase total spending.”
Comment from 31 Mar 2026 in “Universal Claude.md – cut Claude output tokens by 63%” “You're paying this disproportionately-huge amount upfront to save a pittance.”
Comment from 31 Mar 2026 in “Universal Claude.md – cut Claude output tokens by 63%”
Other agent skills people recommend
The four most recommended other agent skills, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Superpowers | 138 | 72 | 66% | 12 | 4 mo7 owners | — |
| Claude Skills | 59 | 26 | 69% | 11 | — | Approved |
| grill-me skill by Matt Pocock | 39 | 2 | 95% | 8 | — | Approved |
| Get Shit Done | 31 | 32 | 49% | 7 | — | — |
| Universal Claude.mdThis page | 10 | 15 | 40% | 1 | — | — |
What people said
All 25 comments that mention it, newest first. Each links to the original comment.
“I got no better results for one or two years now using meta-prompts that are generic and/or from the internet.”
“Your system is great and I do similar”
“forcing short responses can hurt coherence in longer agentic sessions”
“Why does is this ridiculous thing trending on HN?”
“Yeah this seems to be a very bad idea.”
“most of this repo is solving the wrong problem”
“I see no point in this project.”
“I can't see how this gets on the front page? Anyone who interacted with llms for at least a hour, could've figured out to say somethin like "be less verbose"”
“This was ripped apart on Reddit”
“the last line is a good one to have, unless you run a service for other users”
“You're paying this disproportionately-huge amount upfront to save a pittance.”
“efficiency is highest without any instructions”
“I can't calculate how many times this exact section of Claude output let me know that it was doing the wrong thing”
“The benchmark is totally useless. It measures single prompts, and only compares output tokens with no regard for accuracy.”
“Those are some terrible rules. You don't want to be doing that.”
“this is so naive and optimistic”
“For complex tasks this is not a useful prompt.”
“If you're goal is quality outputs, then it is likely too extreme, but there are otherwise useful instructions in this repo to (quantifiably) reduce verbosity.”
“Projects like these tend to feel shortsighted.”
“Considering the 1:5 (input:output) token cost ratio, this could increase total spending.”
“this would, generally, make the model worse”
“I'd rather use the "instruction budget" on the task at hand”
“without comparing before/after performance on some standard LLM benchmark to see if the instructions hurt intelligence”
“the benchmarks here are heavily biased towards single-shot explanatory tasks”
“While I love this set of prompts, I've not seen my clause opus 4.6 give such verbose responses when using Claude code.”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.