What problem does it solve? Setting up a new project with consistent structure for AI-assisted development is repetitive and error-prone. This Skill standardizes project initialization by creating the full directory layout, agent memory/handoff files, spec-driven development documents, Git setup, and optional academic paper-writing scaffolding in one guided workflow. ## Core Features & Use Cases - Three execution modes: Create a new project (Mode A), audit and update an existing project against the template (Mode B), or consult interactively (Mode C). - Standard project skeleton: Generates .agent/ (memory, handoff, workflows, scripts), .spec/ (PLAN/SPEC/TODO/KNOWLEDGE), CLAUDE.md, AGENTS.md, GEMINI.md, .gitignore, and /newplan and /handoff commands, then initializes Git and pushes to GitHub. - Paper project support: When flagged as a paper-writing project, additionally creates LaTeX folders (latex原稿/), a compile.sh template (elsarticle + dvipdfmx), main.tex and section skeletons, refs.bib, reference-verification folders, and numeric traceability templates linked to the paper_writing skill. - Drive archive automation: Determines a Google Drive archive destination and generates sync_to_drive.sh for rsync-based backup of deliverables and records. - Use Case: A researcher starting a new Interspeech paper answers a few prompts and receives a fully initialized Git repository with LaTeX scaffolding, compile script, reference-checking workspace, and Drive sync—ready to write immediately. ## Quick Start Ask the AI to run the make_project skill to initialize a new project, then choose mode 1 and provide the project name and GitHub repository URL when prompted.
