Mathematical Problem-Solving Meets Clean Architecture.
Pólya SDLC Coder prevents AI coding agents from prematurely emitting code by enforcing George Pólya's 1945 mathematical heuristics, Uncle Bob's Clean Architecture seams, and The Mandatory Pause Rule.
Condition(Data) ➔ Find(Unknown) ➔ Proof & Verification
Install Pólya SDLC Coder
Choose your preferred installation pathway: agentic package installation or one-liner shell script.
npx skills add GulajavaMinistudio/polya-sdlc-coder -y
Installs directly into your repository .agents/ directory for GitHub Copilot Workspace and VS Code.
/to-init setup this project
Prompt your AI coding agent with /to-init. This automatically scaffolds AGENTS.md in your project root, configures .agents/rules/ (including PolyaOrchestrator.md), establishes .agents/standards/ (ADR & glossary formats), and initializes .agents/instructions/ with zero data loss. Explore /to-init guide ➔
Prefer an automated terminal script without npm?
The Canonical 6-Phase Lifecycle
Over 90% of AI-generated bugs stem not from syntactic errors, but from premature coding on misunderstood requirements. Pólya SDLC Coder guarantees disciplined progression before writing application code.
Discovery & Spikes
Explore problem feasibility, critique tech debt, and challenge premature premises.
/to-explore, /to-prd
Understand & Specify
Deconstruct Unknown, Data, Condition into DTO contracts and Clean Architecture seams.
/to-spec, /to-clarify
Devise Plan & Gate
Slice into vertical tracer bullets with Plan B contingencies and halt at The Pause Rule.
/to-plan ➔ 🛑 PAUSE
Carry Out & Review
Clean code implementation, 5 SOLID audit, bisection bugfix, and Diátaxis documentation.
/to-code, /to-review
The 15 Modular Agent Skills
Portable markdown specifications executed directly by Claude Code, Copilot, Cursor, Antigravity, and Codex.