Harness Guards

What problem does it solve?

Harness Guards standardizes how safety and governance constraints are injected into plans, phases, and workflow steps so an AI agent follows the right rules at the right moment without relying on generic warnings.

Core Features & Use Cases

• Centralized guard catalog: Defines the exact guard types, syntax, placement, and escalation patterns used across PARA artifacts.
• Auto-scan guard protocol: Scans project “DNA” from rule/skill indices and the Risks & Mitigations table to produce plan-specific, constraint-specific harness guards.
• Sidecar data-only design: Provides references for just-in-time guard generation while keeping guard logic out of the skill itself.
• Use case: When generating a phased plan, the agent loads the guard catalog and scan protocol to insert mandatory and risk-specific harness comments that gate side-effects, enforce checkpoint actions, and protect critical files.

Quick Start

Ask the agent to create a phased plan for your project; it will load Harness Guards and generate context-aware harness comments before writing or modifying plan artifacts.