meta-software-development

Enforces disciplined software engineering rules with first-principles decomposition and execution-based verification.

1|Updated Jun 21, 2026
One-click install
npx skills add https://github.com/tapway/shogun-os --skill meta-software-development-tapway
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: meta-software-development
Source: https://github.com/tapway/shogun-os/tree/main/skills/software-development/meta-software-development
Command: npx skills add https://github.com/tapway/shogun-os --skill meta-software-development-tapway

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Software development tasks often suffer from vague planning, over-engineering, unverified claims of success, and constant mid-task interruptions. This Skill replaces the generic brainstorm-plan-execute flow with a rigorous engineering charter (R1-R11 rules plus a Zero-Pause execution layer) so coding work is decomposed from first principles, verified by actual execution, and shipped continuously without artificial pauses. ## Core Features & Use Cases - First-Principles Decomposition (R1-R11): States root invariants, callers, and failure modes before writing code, with calibrated decisiveness, proportional simplicity, and push-back duty on flawed premises. - Zero-Pause Execution Layer: Maintains continuous momentum with pre-work-only questions, a humanpending.md protocol for human-gated decisions, and parallel multi-role reasoning threads. - Verification by Execution: Defines executable success criteria upfront, writes contract-encoding tests, and tags every claim as (executed), (inspected), or (assumed). - Use Case: When asked to add user authentication to an app, the Skill states assumptions, defines success as "login returns JWT, invalid credentials return 401", executes without pausing, runs the tests, and reports results with calibrated tags. ## Quick Start Ask the agent to build, implement, or ship a software feature and it will automatically load this charter and execute under the META rules.

Frequently Asked Questions about meta-software-development

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I enforce disciplined coding practices with an AI agent?

Load this charter for any software development task and it replaces the default workflow with R1-R11 rules: first-principles decomposition, proportional simplicity, verification by execution, and calibrated reporting. The agent states assumptions upfront and tags every claim as executed, inspected, or assumed.

How to stop an AI coding agent from pausing mid-task?

The Zero-Pause execution layer (ZPR1-ZPR4) eliminates artificial phases and mid-task confirmation requests. Questions are only allowed before work begins, and human-gated decisions are logged to a humanpending.md file while all non-dependent work continues in parallel.

When should I not use this software engineering charter?

Skip it for non-software tasks like research, writing, or operations, and for trivial changes such as typo fixes, one-liners, or simple config edits. Tasks already covered by a more specific skill should use that skill instead.

Does this charter work with existing CLAUDE.md or AGENTS.md files?

Yes, the agent reads project-specific CLAUDE.md or AGENTS.md files first and matches existing conventions by default under rule R11. The charter takes precedence over generic instructions but defers to project-specific technical decisions.

How does the charter handle irreversible changes like database schema migrations?

Rule R10 applies reversibility-weighted verification: local reversible changes ship boldly, but schema changes, production data mutations, and public API modifications require explicit user confirmation. Staging verification is required before production execution.