System Patterns

Architectural Overview

A content- and instruction-centric repository published via Jekyll (GitHub Pages) + Markdown assets. Python scripts provide automation examples; no central runtime service.

Core Components

  • Documentation (Diátaxis) under docs/
  • Instruction sets _instructions/ with applyTo selectors
  • Prompts _prompts/
  • Modes _modes/
  • Automation scripts scripts/
  • Memory Bank memory-bank/ (this directory)

Patterns in Use

| Pattern | Description | Rationale | |———|————-|———–| | Diátaxis Docs | Tutorial/How-to/Reference/Explanation segmentation | Cognitive clarity & onboarding efficiency | | Conventional Commits | Structured commit messages | Automated changelogs & review clarity | | Memory Bank Files | Persistent context across sessions | Overcome stateless AI limitations | | Security Instructions Overlay | OWASP-aligned enforcement | Reduce risk of insecure code suggestions |

Data & State

  • No database: Markdown is the primary state medium.
  • Generated artifacts (results/) are ephemeral and excluded from versioned knowledge unless curated.

Extensibility Strategy

Add new instruction sets via _instructions/ and update Navigation / docs index as needed. Introduce new automation scripts under scripts/ with isolated dependency footprints.

Risks & Mitigations

| Risk | Mitigation | |——|————| | Instruction drift vs code reality | Enforce updates via PR review checklists | | Security regression | Central OWASP instruction file referenced in changes | | Knowledge fragmentation | Mandatory Memory Bank update policy |

Open Questions

  • Should we add automated linting for instruction front matter? (Pending)
  • Standard test harness for scripts? (Planned expansion)