Pre-Delivery QC · Bilingual
AFP five-layer checks before you submit
Pre-Delivery QC · Bilingual
AFP five-layer checks before you submit
Showcase
Description
Perform quality control before a formal document is delivered, submitted, or published, in either Chinese or English. The skill follows an Auto-Flow Prompt (AFP) five-layer workflow: the Trigger Layer identifies the task and creates a record; the Planning Layer generates review items and an execution plan from the routing table; the Execution Layer adaptively reviews each item, including factual claims, certainty levels, task omissions, numerical standards, logic and structure, compliance and privacy, practical usability, audience mismatch, proper nouns, timeliness, layout and formatting, figures, links, attachments, collaboration traces, and more. The Supervision Layer runs a “check → challenge → verify → correct → retest” self-review loop and flags items requiring human intervention. The Integration Layer removes duplicates, assigns priority levels, resolves conflicts across items, and consolidates the results into a report. The final output includes a P0/P1/P2 issue list with each item’s location, basis, and suggested action, along with an overall verdict: “Releasable / Releasable after revision / Not releasable.” The skill automatically detects the document’s language and produces the report in that language. It reports only: it does not modify files or automatically send or publish them. Final release authority and responsibility remain with a human.
Related Skills
View all
WriteAIGC Reduction & Rewrite v7.0
📚 Academic Paper AIGC Reduction and Quality-Preserving Rewriting Expert v7.0 Designed for graduate students, researchers, and paper authors, this academic text optimization skill operates on a core closed loop of 'source control → process correction → result verification → reverse self-check → iterative re-check'. It systematically diagnoses and optimizes issues such as templated expressions, mechanical logic, vague content, inaccurate terminology, and style inconsistencies while preserving original meaning, technical terms, data, and core conclusions. 🔍📝 🌟 Core Capabilities 🔬 Multi-layer Risk Diagnosis Covers 10 types of universal text fingerprints and assists in identifying common expression patterns of models such as ChatGPT, Claude, DeepSeek, and Wenxin Yiyan. 🧠 Deep Semantic Restructuring Goes beyond synonym replacement to rebuild more natural and in-depth academic reasoning by adjusting proposition expression, information order, argument approach, and evidence organization. ✍️ Quality-Preserving Rewriting Comprehensively applies 13 sentence transformation strategies and 20 methods for cleaning high-frequency templated expressions, improving mechanical sentence structures, repetitive connectors, and overly rigid formatting. 📊 Full-text Structure Diagnosis Through macro-cycle and five key triangles, checks whether research problems, theory, literature review, methods, results, conclusions, and innovation form a complete closed loop. 🧩 Fine-grained Section Adaptation Develops differentiated diagnostic and rewriting strategies for abstract, introduction, literature review, research methods, results, discussion, and conclusion respectively. 🌐 Cross-language Risk Scanning Assists in identifying translationese, passive voice stacking, long sentence nesting, and mixed Chinese-English formatting abnormalities to make Chinese academic expression more natural and accurate. 🔄 Reverse Self-check Loop After rewriting, re-verifies from three aspects: technique distribution, new text fingerprints, and information integrity, to avoid 'becoming more templated' or losing key content. 🛡️ Academic Integrity Protection Does not fabricate literature, data, cases, or policy evidence; separately marks information requiring author verification and reminds users to honestly disclose AI usage. 🎯 Use Cases ✅ Single paragraph or partial section optimization ✅ Targeted modification of marked paragraphs from inspection reports ✅ Polishing of abstract, introduction, literature review, discussion, and conclusion ✅ Full-text AIGC risk feature diagnosis ✅ Language and structure adaptation for target journals ✅ Pre-submission quality review and consistency check 📦 Final Deliverables 📄 Quality-preserving rewritten text 🔎 Risk and issue diagnosis report 🛠️ Rewriting strategy and technique description ✅ Reverse self-check and information integrity report 💡 Items requiring author verification and subsequent revision suggestions 🎓 Original meaning preserved · Logic intact · No fabricated data · Academic quality maintained ⚠️ This skill aims to improve academic expression quality and reduce text risk features. It does not guarantee passage through any specific detection platform or achieving a particular detection score.
WriteSocial Content Precheck
Before publishing on WeChat Channels, WeChat Official Accounts, Douyin, Xiaohongshu, Kuaishou, Bilibili, Zhihu, Weibo, or other platforms, get a precheck covering compliance risks, audience response, and reach potential. Whether it is a short-video script, image-and-text post, WeChat Official Account article, Moments copy, livestream script, or direct-message script, the precheck starts with the original text, identifies the platform, category, target audience, and publishing goal, then flags risks such as performance promises, absolute claims, restricted areas in regulated industries, and inducements to move users off-platform. The precheck uses traceable, up-to-date rules to flag high-, medium-, and low-risk wording line by line, explain the basis for each judgment, and provide ready-to-use replacement copy that preserves the original tone and rhythm. It also highlights differences in review standards across platforms. For the target audience, it simulates reactions in the first three seconds or at the start of reading, generates authentic-sounding comments showing support, skepticism, and backlash, and identifies the sentence most likely to attract criticism. For reach evaluation, the Skill provides explainable scores based on the hook, emotion, information gain, identity, quotability, likelihood of being read or watched to completion, and path to action. It gives platform fit, probability ranges, and the highest-priority changes. The final delivery includes an in-chat overview and a full report, helping you quickly decide whether to publish as is, publish after revisions, or rewrite—and provides a clean version of the copy for retesting.
Chinese Jargon & SVO Review
Rewrites Chinese materials that “sound professional but leave readers unsure who needs to do what” into language readers can understand, act on, and verify directly. It applies to reports, proposals, notices, articles, product documentation, project updates, policy training, customer service content, official WeChat articles, and AI-generated text, with a focus on Chinese jargon, vague slogans, abstract nouns, and writer-centered framing. Without changing facts, quantities, dates, names, scope, the strength of conclusions, or the boundaries of the evidence, this Skill turns difficult-to-act-on phrases such as “empower,” “strengthen management,” “drive implementation,” “close the loop,” and “align through collaboration” into more specific people, actions, objects, conditions, and outcomes. It also checks whether key sentences clearly state the subject, action, and object, so readers know who is responsible for what, what is affected, and how to determine whether the work is complete. When necessary, it retains specialized terms from regulations, standards, and products while adding explanations in plain language. The output typically includes a ready-to-use rewrite, the key changes that affect understanding, and any people responsible, timelines, data, evidence, or specialized definitions that cannot be confirmed from the original. It does not invent information or rewrite suggestions, records, or system statuses as facts that are completed, resolved, or verified. It is suited to writers, editors, and business teams who want to improve the readability, accuracy, and actionability of Chinese materials.
Information
- Version
- v5
- Last updated
- Runtime credits
- Usage-based
- Models
- Auto
