tailored-resume-generator

More from touricks/fanshi_personal_skills

• humanizer-zh

  去除文本中的 AI 生成痕迹。适用于编辑或审阅文本，使其听起来更自然、更像人类书写。基于维基百科的\"AI 写作特征\"综合指南。检测并修复以下模式：夸大的象征意义、宣传性语言、以 -ing 结尾的肤浅分析、模糊的归因、破折号过度使用、三段式法则、AI 词汇、否定式排比、过多的连接性短语。

  6

• humanizer

  Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive \"Signs of AI writing\" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, negative parallelisms, and excessive conjunctive phrases.\n\nv3.0 adds: Academic writing patterns (literature reviews, paper critiques, research summaries) that evade general-purpose detection but trigger GPTZero and similar classifiers. Covers catalog-style lit reviews, over-clean categorization, uniform confidence, missing first-person engagement, and template-parallel paragraph structures.\n\nv3.1 adds: Second-pass patterns discovered after applying v3.0 and re-scanning with GPTZero (still 99% AI). Covers exhaustive technical description, formulaic first-person insertions, em dash density in academic LaTeX, N-camps reframing trap, and burstiness/information-density uniformity. These are \"second-generation\" AI tells \u2014 they appear in text that has already been humanized once but still triggers classifiers.\n\nv3.2 adds: THE PERPLEXITY CEILING \u2014 the fundamental discovery that pattern-level editing of AI text has a hard ceiling. GPTZero uses statistical models that detect the probability distribution of token sequences, not individual sentences. \"AI edits AI\" retains the statistical fingerprint regardless of surface changes. This version adds a PROCESS-LEVEL strategy: human-first drafting, perplexity injection, structural noise, and a collaborative workflow where AI assists human writing rather than the reverse.\n\nv3.3 adds: GUIDED INTERVIEW MODE \u2014 an interactive Q&A workflow that uses AskUserQuestion to extract the human's genuine thoughts about a paper before any writing begins. The AI acts as a structured interviewer across 4 rounds (first impression, technical engagement, critical analysis, personal connection), then assembles the human's own words into a draft. This is the most reliable method for producing text that passes statistical classifiers, because the core token sequences originate from the human.

  5

• routing

  Scan global skills against project context and produce a tool routing report for developer review. Use when initializing a project, onboarding new tools, or when the developer says "routing", "scan tools", "which skills do I need", "update tool inventory". Produces docs/tool-routing-report.md and STOPS for human approval before any skills are moved.

  1

• call-codex

  Ask OpenAI Codex CLI for a second opinion, critique, or analysis from within a Claude Code session. Use when the user says "ask codex", "call codex", "get codex's opinion", "second opinion from codex", "codex review", or wants to consult Codex on code, architecture, or any technical question. Requires codex-cli installed (`codex` binary available in PATH).

  1

• sentinel-loop

  Propose a Ralph Loop iteration based on current project state. Reads progress.yaml and PRD.md to identify the next task, builds a structured prompt with completion condition, and saves a copy-pasteable /ralph-loop:ralph-loop command. Use when "sentinel-loop", "iterate", "start working", "what's next", or after /boundary completes.

  1