Mnemonica Core - AI Skill

When to use

Activate this skill when:

• User mentions mnemonica, define(), lookupTyped, TypeRegistry
• User asks about prototype chain inheritance in JavaScript/TypeScript
• User is modifying files in core/src/ or core/test/
• User asks about tactica-generated types or .tactica/ directory
• Task involves: type definitions, hooks, async constructors, error handling

High-priority checklist: Adding a new type

When the task involves adding or modifying a type definition:

1. Use type for data, interface for behavior contracts

   • Instance data → type MyTypeData = { ... }
   • Constructor contract → interface MyTypeConstructor { ... }
   • See rules-skill/type-system.md

2. Check TypeRegistry augmentation (if using tactica)

   • Does .tactica/types.ts need updating?
   • Is lookupTyped('MyType') properly typed?
   • See rules-skill/lookup-typed.md

3. Test both success and error paths

   • Mocha test in test/ for runtime behavior
   • Jest test in test-jest/ for type coverage
   • Error path: test ALREADY_DECLARED, WRONG_MODIFICATION_PATTERN
   • See rules-skill/testing.md

4. Run coverage before completing

   • npm run test:cov (Mocha + build)
   • npm run test:jest:cov (Jest on TypeScript source)

Non-discoverable rules (must be in SKILL.md)

• type vs interface rule — not enforced by lint, not in code
• 100% coverage requirement — not discoverable from jest.config.js alone
• Space before function parens — convention, not linted
• Error constructor names are String objects — runtime quirk
• Jest tests must mirror Mocha patterns from test/environment.js — not in code

Build Commands

npm run build          # Full build with linting
npm run test:cov       # Mocha tests with coverage (includes build:all)
npm run test:jest:cov  # Jest tests with coverage
npm run watch          # Watch mode

Rule Reference

Read individual rule files for detailed explanations and code examples:

Type System

• rules-skill/type-system.md — Proto<P,T>, TypeConstructor, type vs interface

Patterns

• rules-skill/define-patterns.md — define() usage, subtypes, config options
• rules-skill/lookup-typed.md — TypeRegistry, lookupTyped, tactica integration
• rules-skill/async-constructors.md — async constructors, awaitReturn, chaining

Architecture

• rules-skill/proxy-architecture.md — TypeProxy, InstanceCreator, Mnemosyne pipeline
• rules-skill/instance-methods.md — extract(), fork(), parent(), internal props
• rules-skill/hooks.md — preCreation, postCreation, creationError

Philosophy & Design

• rules-skill/philosophy.md — HoTT concepts (Univalence, Path Types, Higher Inductive Types) applied to mnemonica's self-reflection model
• rules-skill/ecosystem.md — PACT framework: personas, collaboration modes, integration points

Quality

• rules-skill/testing.md — 100% coverage, Mocha + Jest, error path patterns
• rules-skill/code-style.md — tabs, spacing, type vs interface
• rules-skill/error-system.md — BASE_MNEMONICA_ERROR, constructError, stack cleaning

Contributing

• rules-skill/contributing.md — behavioral guidelines: think before coding, simplicity, surgical changes, goal-driven execution

Contributing

This file covers usage of mnemonica only. If you are modifying the library
itself, read `AGENTS.md` (or `.ai/AGENTS.md`
for framework-agnostic rules).

External Resources

• Repository: https://github.com/mythographica/mnemonica
• Main README: README.md
• Contributor Guidelines: AGENTS.md