Personalized Civil Service Essay Coach

Problem

Civil service essay practice often breaks down into disconnected steps: finding past papers, preparing a printable or on-screen answer sheet, judging the draft, and tracking improvement over time. Candidates may have questions, notes, and scores scattered across chats, PDFs, and local files, making it hard to compare scoring points, recurring weaknesses, and progress across different exams. This skill turns that loop into a repeatable workflow: fetch papers from the Fenbi question bank, generate local study materials, submit answers, review them against scoring points, and write the results back into a long-term profile.

How It Works

The main path is initialize → login → download → answer → review → profile update. First-run setup covers model detection, runtime restoration, license check, workspace creation, and dependency preparation. Afterwards, intents such as downloading a paper set, browsing the question bank, or asking for grading can trigger the corresponding step. The skill can generate Word paper packs, JSON data, and a computer-based answer sheet, supporting both printed and on-screen practice. During review, it focuses on scoring points to show which sentences earn marks and which lose them, then stores the outcome in user_data/profile.json. Over time, this builds a six-dimension ability profile, answer records, and historical trends for cross-paper comparison.

Boundaries

It depends on the DeepSeek model family, a network connection, and a Fenbi account login; license checks and question bank downloads are affected by external conditions. Word documents are better suited for printing or archiving, and some links may only work on the local machine. For daily answering, use the computer-based answer sheet. Because papers, answers, and profile data are written to the local workspace, keep Shenlun practice in a dedicated folder and avoid overwriting user_data/ or the local question bank during updates.