.harness

Imported from yasuguerra/plaud-own (.harness/AGENTS.md). Install upstream with npx skills add yasuguerra/plaud-own --skill .harness. Copyright stays with the author.

HARNESS ENGINEERING SYSTEM — AGENTS.md

Version 3.0

PURPOSE

This harness implements a multi-role software engineering system designed to produce production-ready code with maximum architectural integrity, security, maintainability, and alignment with business objectives.

HOW TO LOAD THIS HARNESS

This file is the INDEX. It does not contain full roles. Each role lives in its own file to preserve the context window.

Only load the active role at any given moment:

Always load alongside the active role:

• project_context.md — technical context of the project
• conflict_resolution.md — decision hierarchy

CORE PRINCIPLES (inherited by all roles)

1. Customer Value First — the customer matters more than technical elegance
2. Security By Default — never as an afterthought
3. Simplicity Over Complexity — the simplest solution that works
4. Reuse Before Creation — look before creating
5. Architecture Before Implementation — plan before coding
6. Verification Before Approval — test before approving
7. Eliminate Before Adding — remove before adding
8. Long-Term Maintainability Over Short-Term Speed — the future matters

GOLDEN RULE

Before creating new code, evaluate in order:

1. Can it be solved with configuration?
2. Can it be solved by reusing existing code?
3. Can it be solved by refactoring existing code?
4. → Only then: create new code

New code is the last resort.

AVAILABLE WORKFLOWS

PERSISTENT MEMORY STRUCTURE

.harness/
├── AGENTS.md                  ← this file (index)
├── project_context.md         ← technical context of the project
├── conflict_resolution.md     ← veto and decision hierarchy
│
├── roles/                     ← one file per role
├── workflows/                 ← one file per workflow type
│
├── architecture/
│   ├── vision.md
│   ├── tech_stack.md
│   ├── coding_standards.md
│   ├── api_conventions.md
│   ├── database_conventions.md
│   └── decision_log.md
│
├── contracts/
│   └── feature_contract.md
│
├── adrs/
│   └── ADR-XXX.md
│
└── audits/
    ├── security_report.md
    ├── entropy_report.md
    └── release_report.md

FINAL RULE

The goal is not to generate the largest amount of code. The goal is to create the simplest, most secure, and maintainable solution that satisfies the business objective.