Skip to content

WARP.md

Latest commit

 

History

History
43 lines (33 loc) · 1.37 KB

WARP.md

File metadata and controls

43 lines (33 loc) · 1.37 KB

llmstrip — context for Warp AI

This repo is a tool that removes LLM-isms from text and code. It works as:

  • a Claude Code skill (/llmstrip)
  • a Cursor rule (.cursor/rules/llmstrip.mdc)
  • an OpenCode prompt
  • a Rust CLI (echo "text" | llmstrip)

Key files

File Purpose
prompts/claude-code.md Claude Code skill — copy to ~/.claude/skills/llmstrip/SKILL.md
prompts/cursor.mdc Cursor rule — copy to .cursor/rules/llmstrip.mdc
prompts/opencode.md OpenCode prompt
prompts/system-prompt.md Universal system prompt for any LLM
rules/ 9 rule files, 137 patterns with before/after examples
cli/ Rust CLI source

Install (quick)

make install-skill   # Claude Code
make install-cursor  # Cursor
cargo install --path cli/  # CLI binary

What it detects

Text: banned words (leveraging, comprehensive, robust, seamlessly...), sycophantic openers, chatbot closers, hedging, rule of three, em-dash overuse.

Code: tautological comments, section headers, bare TODOs, Manager/Handler/Helper suffixes, type-in-name variables, LLM docstring openers, step-numbered comments.

Rules format

Each file in rules/ is markdown with this structure per pattern:

  • Pattern name (sentence case heading)
  • 1-2 sentences on why LLMs do it
  • LLM: before example
  • Human: after example
  • Rule: one-line fix