9S

91160/skills

Developer tools

This page was translated automatically. Technical details may differ from the original. View the original

21 stars Quality 55 Trend 14

Health160 AI Skills Hub — AI skill set covering full-chain R&D scenarios including requirements analysis, project initialization, coding specifications, code scanning, and test case generation, supporting standalone or combined calls.

Overview

— A complete AI coding workflow specification tool. This framework supports AI-Coding for new projects, old projects + new requirements, bug fixes, and technical optimization scenarios, providing high-quality, efficient workflows and complete execution steps. Supports multiple AI coding tools (claude/codex/cursor, etc.). 1. Write the AGENTS.md (G series core rules + numbering flow specification) into the project root directory, with paths automatically parsed. 2. Create instruction files for each AI tool symlink 3. Output the initialization status report NY-SDD-Workflow is a set that defines the entire process from project initiation to business development: - :Before AI writes code, it must pass 6 checks (project initialization, requirements documentation, solution review, skill installation, task positioning, iron rule compliance) - :5 types of uncertain scenarios automatically pause to await user judgment - :Automatically matches company-level coding specification skills according to the tech stack, enforces logs + coding-skill/audit-skill dual-column status tracking - :.docs/ directory automatically scans and recognizes, coding must be based on documentation - :module sorting + Dependency analysis + subtask splitting + continuous/itemized confirmation execution mode - :P three trigger sources: RD change / technical documentation change / specs change, unified cascading processing - : New project 6-dimensional audit, single-function 4-dimensional audit - : 10 coding protocols (C-01~C-10) + 8 audit protocols (S-01~S-08 including UI restore + functional integrity) + 6 self-test items (T-01~T-06) - : Passthrough / fast / standard three-channel, automatic routing by change type - : Interface documentation / audit report / self-test report / execution report / change records, automatically added by module chapter - : One rule file automatically adapts to 8 AI coding tools Break rules into **, and AI only loads the rules needed for the current stage at a time, avoiding attention dilution and token waste. - (AGENTS.md Always load): G0 Dialogue initialization + G1 Code writing access control + G2 Stop signal + G3 Uncovered scenario back-up - (phase-*.md load on demand): §1 Project Initiation / §2 Requirements and Design / §3 Encoding Change Channel / §4 Archiving - At the top of each §N.N chapter, there is a "Process Declaration" reference block (phase/step/prev/next/gate/blocking), which AI mechanically jumps by field, no need to reason from the main text AGENTS.md Use the {SKILL_DIR} placeholder to reference the stage file in the template.

Overview (machine translated)

SDD Workflow — A comprehensive AI coding workflow specification tool. This framework supports AI-coding for new projects, old projects + new requirements, bug fixes, and technical optimization scenarios, providing high-quality, efficient workflows and complete execution steps. Supports multiple AI coding tools (claude/codex/cursor, etc.).

One command to install, unifying team AI coding standards.

Installation

Method 2: Manually install via Git Clone

After installation

When the AI says “Start workflow” or “Install SDD” in an AI conversation, the AI will automatically:

  1. Write AGENTS.md (G series core rules + numbering process specification) into the project root directory, and the path will be automatically parsed
  2. Create the command files for each AI tool called symlink
  3. Output the initialization status report

What is it?

NY-SDD-Workflow is a set of AI coding workflow rules that define the entire process from project initiation to business development:

  • G1 Code Writing Access Control: AI must pass 6 checks before coding (project initialization, requirements documentation, solution review, skill installation, task positioning, strict compliance)
  • G2 Stop Signal: Automatically pauses 5 types of uncertain scenarios awaiting user decision
  • Skill Routing: Automatically matches company-level coding specification Skills according to the technology stack, enforcing dual-column status tracking with logs + coding-skill/audit-skill
  • Technical Documentation: .docs/ directory is automatically scanned and identified; encoding must refer to the document
  • Task Splitting: Module sorting + dependency analysis + subtask splitting + continuous/itemized confirmation execution mode
  • Requirement Changes:P Three trigger sources: RD changes, technical documentation changes, and specs changes, all handled in a unified cascade
  • PRD audit: New project 6-dimensional audit, single-function 4-dimensional audit
  • Coding Protocols: 10 coding protocols (C-01~C-10) + 8 auditing protocols (S-01~S-08 including UI restore + functional integrity) + 6 self-test items (T-01~T-06)
  • Change Channels: Three-channel (pass-through/fast/standard), automatic routing based on change type
  • Delivery Documents: interface documentation / audit report / self-test report / execution report / change records, automatically added by module chapter
  • Multi-tool Compatibility: One rule file automatically adapts to 8 AI coding tools

Architecture: Dynamic loading + process declaration header + front-end and back-end context independent

Design Approach

Break rules into core + phase files, with AI loading only the rules needed for the current stage at a time, avoiding attention dilution and token waste.

v1.0 adopts two numbering systems + process declaration header mechanism:

  • G Series (AGENTS.

This page was translated automatically. Technical details may differ from the original.

Recommended Tools

Try a different keyword or remove a filter.

Install

npx skillfish add 91160/skills
New to MCP? Start with the mcpskill.org guide →