agentleFS
NovusEdgeNovusEdge/curt/skills/anti-slop/SKILL.md

anti-slop

Skill1 starsChanged 34 days agoMITView on GitHub

Use when writing or editing prose that feels padded, formulaic, promotional, or unlike its author, including technical docs, personal blogs, portfolio pages, comments, and PR descriptions. Use anti-slop-code for code structure.

31 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use research and handoff.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • NVIDIASkill

    research

    Before designing a feature, research how comparable libraries solve the problem, check best practices, known pitfalls, and upstream changes. Use when starting a new feature, investigating an unfamiliar area, or the user asks to research something.

    Also in 3 projects that use this

  • MicrosoftSkill

    handoff

    Compact the current conversation into a handoff document for another agent to pick up.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 30 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Default to the smallest useful edit.

    line 11
  2. Distinctive:

    Use the supplied facts for a wording-only edit.

    line 23
  3. Distinctive:

    Treat these as contextual signals, not forbidden strings.

    line 37
  4. Distinctive:

    Keep useful jokes, profanity, contractions, opinions, and varied sentence lengths already present in the author's work.

    line 46
  5. Distinctive:

    Read the revision against the original.

    line 63

“Common” is measured: kept by at least 6 in 10 of the 30 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Anti-slop

Improve the reader's understanding while keeping the author's meaning and voice. Default to the smallest useful edit. A successful pass can leave a paragraph unchanged.

Work from the text

Establish the audience, purpose, and requested action. A review returns findings; an edit changes the requested material. Read enough surrounding text to understand the argument and distinguish the author's habits from an isolated awkward line.

Preserve facts, qualifications, examples, links, technical identifiers, and the distinction between planned, implemented, and tested behavior. Keep the reasons and transitions that connect the argument. A wording pass is not a summary.

Use the supplied facts for a wording-only edit. Separate a suspected factual problem from the editorial fix: flag it or check the relevant source when that is in scope. Do not silently strengthen, weaken, or resolve the claim to make the sentence cleaner. Never invent an anecdote, motive, feeling, or first-person experience to make writing sound human.

Edit the passage, not its vocabulary

First identify what the sentence contributes. Remove repeated information or ceremony, then repair only the connections the cut affects.

Watch for passages that stage a revelation, announce honesty, praise the author's own skepticism, ask a question merely to deliver a punchline, or close every section with a slogan. Keep the actual correction, evidence, or implication. Treat these as contextual signals, not forbidden strings. A real change of mind can be essential to an essay. Retain what the author previously believed and what changed it when the text supplies those facts. An opener such as "there goes a line I'd been leaning on" alone supplies neither; cut it without inventing the missing belief.

Replace an empty promise with the supported behavior and its limits. If those facts are missing, flag the gap. Do not substitute a quieter unsupported promise.

Keep useful jokes, profanity, contractions, opinions, and varied sentence lengths already present in the author's work. Do not manufacture casualness with new slang, deliberate mistakes, clipped fragments, or a stock cynical persona.

For technical instructions, make actors, conditions, sequence, and consequences clear. Split a sentence when its structure obscures those relationships; keep related clauses together when that makes the relationship easier to follow. For blogs and portfolio pages, retain the argument and the reasons the author cares. A page about an unfinished project must remain clear about its status.

Word lists and lint matches suggest places to look. They do not establish authorship or prove a defect. Passive voice, contrast, ordinary adjectives, headings, apologies, and lists can all serve a real purpose. There is no sentence length cap, mandatory STE grammar, or target reduction percentage.

Compare before finishing

Read the revision against the original. Check for lost examples and arguments, changed uncertainty or numbers, missing conditions, newly invented experience, and a replacement voice that is uniformly terse or polished. Restore anything meaningful lost merely for brevity. Read the paragraph as a whole for rhythm and logical connections.

Return the requested edit. For a review, cite the passage, its reader-facing problem, and a specific correction. Report only material unresolved questions and verification limits. Do not add a ceremonial account of the editing process.

For a difficult passage, consult structural-patterns.md. Other references cover word choice, technical clarity, and responding to corrections.

<!-- anti-slop: ignore-file (this file discusses lint patterns) -->

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.