v1truv1us/ai-eng-system

text-cleanup

Remove AI-generated filler, repetition, inflated language, and redundant comments while preserving meaning.

First seen Jan 25, 2026

Installation

$ npx skills add v1truv1us/ai-eng-system --skill text-cleanup

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from v1truv1us/ai-eng-system · top by installs.

npx skills add v1truv1us/ai-eng-system

Browse all from v1truv1us/ai-eng-system

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Also listed on

Alternate registries and mirrors of this skill.

Repository health

Stars 8
License LICENSE
Default branch main
Open issues 0
Status Active

Skill metadata

Parsed from SKILL.md frontmatter.

Version2.0.0
More metadata
category
user-invoked
version
2.0.0
tags
text-cleanup, slop-removal

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 1,771 B
  • docs SUMMARY.md 127 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 21 installs

SKILL.md

Default output: return the cleaned text only. For file edits, report changed files and verification in at most three bullets.

Text cleanup

Cut words that do not add facts, constraints, decisions, or useful context.

Remove

  • Conversational openers and closers.
  • Fake urgency, rewards, penalties, challenges, and fictional credentials.
  • Generic instructions such as "be thorough," "think step by step," or "work systematically."
  • Confidence self-ratings unless backed by a defined metric.
  • Repeated conclusions, transitions, and restated headings.
  • Self-evident comments that repeat code.
  • Vague claims such as "experts say" without a source.
  • Inflated words when a plain word is accurate.

Preserve

  • Technical constraints, numbers, formulas, and acceptance criteria.
  • Error cases, security warnings, and compatibility limits.
  • Decisions and their rationale.
  • Examples that clarify format or edge behavior.
  • Required schemas and domain terminology.

Rewrite rules

  • Lead with the result.
  • Prefer active voice and concrete nouns.
  • Replace vague quality words with measurable requirements.
  • Keep one idea per sentence.
  • Use the minimum headings and bullets needed for scanning.
  • Do not shorten text when brevity would remove necessary precision.

Code cleanup

When cleaning branch code, limit edits to introduced slop: redundant comments, unnecessary wrappers, abnormal defensive handling, type escapes, and avoidable nesting. Preserve behavior and run affected checks.