← ClaudeAtlas

removing-llm-tropeslisted

Use when writing or rewriting text for a reader outside the session: commit messages, PR titles and descriptions, documents, issues, tickets, and messages. Also use when the user asks to humanize text, remove AI slop, or unglaze writing. Writes the draft to the explanation guide of the writing plugin, then runs the upstream Writing Whip ban list and trope checklist as a separate post-generation audit and rewrite.
rgant/claude-plugins · ★ 0 · AI & Automation · score 75
Install: claude install-skill rgant/claude-plugins
# Removing LLM Tropes This skill adapts the Writing Whip and the trope list from [tropes.fyi](https://tropes.fyi) by [ossama.is](https://ossama.is). Both are under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). The upstream text is in two gists: [writing whip](https://gist.github.com/ossa-ma/dae6f9571534f3fbd1266a384be00e11) and [tropes](https://gist.github.com/ossa-ma/f3baa9d25154c33095e22272c631f5a1). This copy starts from the `whip` skill in [rednafi/dotfiles](https://github.com/rednafi/dotfiles/tree/main/dot_agents/skills/whip). The `metadata.source` link pins the copied commit. The upstream skill is for blog articles. This copy has local edits for technical writing in a codebase. ## Scope Apply this skill to text for a reader outside the session. The user's own instructions and `${CLAUDE_PLUGIN_ROOT}/standards.md` control replies to the user in the session, and code comments and docstrings. ## Conflicts If a rule in this skill conflicts with the user's own instructions or with `${CLAUDE_PLUGIN_ROOT}/standards.md`, obey the instructions or the standards. Then do these steps: 1. Tell the user in the session, in one line. Name both rules and the rule that you obeyed. 2. Add one line to `${CLAUDE_PLUGIN_DATA}/removing-llm-tropes/CONFLICTS.md`. Create the file if it is absent. Give the date, the skill rule, the rule that you obeyed, and the type of text. 3. If that `CONFLICTS.md` already has a line for the same skill rule, ask the user for an ed