Analyze technical documentation prose defects and generate prioritized improvement lists.
npx skills add https://github.com/anthony-chaudhary/fak --skill appeal-score
Please help me install this Agent Skill. Skill: appeal-score Source: https://github.com/anthony-chaudhary/fak/tree/main/.claude/skills/appeal-score Command: npx skills add https://github.com/anthony-chaudhary/fak --skill appeal-score
This Skill addresses the issue of machine-generated prose that feels robotic, repetitive, and difficult for human readers to engage with, ensuring your documentation sounds authentic and trustworthy.
Run the appeal score tool on the README file to identify prose defects and generate a prioritized list of human-voice improvements.
High-intent search queries and answers about installing and using this skill.
To make technical documentation sound less robotic, you can run a prose scorecard that analyzes text for specific LLM-tells like em-dash floods, overlong sentences, and scaffolding phrases. This process generates a prioritized list of human-voice improvements.
Common LLM-tells in README files include em-dash floods, overlong sentences, and repetitive LLM-scaffolding phrases. You can audit these prose defects using a metric-driven scorecard that identifies readability issues while maintaining strict factual and link integrity.
You can improve the readability of technical guides without altering technical accuracy by applying a debt burndown workflow. This approach provides a repeatable, metric-driven process to reduce prose defects while strictly maintaining the underlying claims and factual integrity.
Yes, you need Python installed to execute the scorecard logic for readability audits. The Python environment is required to analyze reader-facing prose, identify machine-generated defects, and generate actionable improvement lists.
You should audit your documentation for machine-generated prose defects before a release to ensure the content lands effectively with human readers and answer engines. This guarantees your technical guides sound authentic and trustworthy without robotic repetition.
encero-systemsReview Markdown documentation for truthfulness, prose quality, and RFC leakage.
atelier-fashionAudit a codebase with four parallel agents and produce a prioritized health report.
prismaRemove artificial hard line breaks from Markdown prose paragraphs.
jandragsbaekDetect AI-writing patterns and rewrite text into natural prose.
