Skip to content

CLAUDE.md

Latest commit

 

History

History
130 lines (94 loc) · 5.74 KB

CLAUDE.md

File metadata and controls

130 lines (94 loc) · 5.74 KB
Raw

Claude Code Instructions for Eyas

Active TODOs

See TODO.md for the current punch list of outstanding work. When working in an area touched by an open item there, surface it and offer to address it — don't silently ignore it. Once you're confident an item is fully resolved, delete its line rather than marking it done. Whenever you touch TODO.md, also normalize its formatting: under each of the two headers, plain - bullet syntax, one item per line, no stray blank lines. TODO.md always has this baseline present, even when both sections are empty:

# TODO

-


# Future

-

Project Overview

Eyas — a QA/testing desktop application built with Electron, Vue 3, and TypeScript. Eyas simplifies hands-on testing for web applications, allowing teams to quickly bundle and test changes in any state before release. Built with electron-vite for fast development iteration.

Build & Test Commands

# Local development (Electron + dev server)
npm run dev

# Type checking
npm run type-check

# Linting
npm run lint

# Run all tests
npm run check

# Test Suites (targeted)
npm run test:electron:unit       # Electron core tests (vitest)
npm run test:interface:unit      # Vue interface tests (vitest)
npm run test:demo:unit           # Demo app tests (vitest)
npm run test:electron:e2e        # End-to-end tests (Playwright)

# Compilation
npm run compile:build            # Build Electron app
npm run compile:runner           # Compile platform runners
npm run compile:mac              # Full macOS build + sign + notarize
npm run compile:installer        # Platform-specific installer build

# Code auditing
npm run audit-dead-code          # Find unused code

Code Discovery

Prefer the codebase-memory-mcp tools for code exploration:

  • search_graph(name_pattern/label/qn_pattern) — Find functions/classes/routes by name
  • trace_path(function_name, mode=calls|data_flow|cross_service) — Trace call chains or data flow
  • get_code_snippet(qualified_name) — Get exact source with precise line ranges
  • query_graph(cypher_query) — Complex graph pattern queries
  • search_code(pattern) — Text search augmented by graph data

Only fall back to Grep/Glob for text in non-code files (configs, docs) or when the graph doesn't have an answer. Project name for codebase-memory: Users-erichigginson-repos-eyas.

Detailed Standards & Patterns

See AGENTS.md (root) for comprehensive engineering standards covering:

  • Core philosophy (Strategic Narrative Planning, "Stop. Think. Plan. Verify. Only then, Code.")
  • BDD (Behavior-Driven Development) approach and mandatory testing standards
  • TypeScript, DRY, linting, and type safety conventions
  • Refactoring, efficiency tiers, and verification gates
  • Technology stack, patterns, and operational workflows

Module-Specific Standards

Refer to module-level AGENTS.md files for area-specific patterns and constraints:

  • src/types/AGENTS.md — Type registry, interface conventions, and semantic naming
  • src/eyas-core/AGENTS.md — Electron core process, IPC patterns, and main-thread standards
  • src/eyas-interface/AGENTS.md — Vue 3 interface, component patterns, and state management
  • src/eyas-interface/app/src/AGENTS.md — Interface-specific organizational patterns
  • src/eyas-interface/app/src/components/AGENTS.md — Component standards and testing patterns
  • src/scripts/AGENTS.md — Script execution and tooling patterns
  • tests/AGENTS.md — Test organization, frameworks, and verification strategies

Claude Code Skills

This project includes 7 Eyas-specific Claude Code skills in .claude/skills/. They auto-invoke based on task context and provide domain-specific guidance.

Development & Implementation

  • /vue-interface-standards — Vue 3 components and Composition API
  • /electron-core-standards — Electron main process architecture and IPC patterns

Testing & Quality

  • /testing-standards — Vitest execution, mocking, and test organization
  • /type-registry-standards — Type definitions and registry organization

Platform & Tooling

  • /electron-e2e-testing — Playwright E2E test strategies and synchronization
  • /active-test-content-gating — Post-mortem on visibility gating patterns

Generic, project-agnostic skills (planning, BDD philosophy, post-mortems, refactoring patterns, TypeScript gotchas, feature flagging, config/directory auditing) now live at the user level (~/.claude/skills/) and apply automatically without needing an entry here — see .claude/skills/README.md for the full breakdown of what moved.

See .claude/skills/README.md for the complete skills index and auto-invocation triggers.

Common Tasks

  • Adding a Vue component: See /vue-interface-standards skill and src/eyas-interface/app/src/components/AGENTS.md
  • Modifying Electron core: See /electron-core-standards skill and src/eyas-core/AGENTS.md
  • Writing tests: See /testing-standards skill

Development Environment

  • Node version: Check .nvmrc for required version
  • Dependencies: Run npm install after cloning
  • TypeScript: npm run type-check for compile verification
  • Linting: npm run lint for style enforcement
  • Pre-commit: Configured hooks run linting and type checks on staged files

Workflow Checklist

  1. Plan: Write Strategic Narrative (see /cognitive-pre-processor)
  2. BDD: Write test case headers (it.todo) before implementation (see /bdd-planning)
  3. Test: Write tests, verify they fail (Red phase)
  4. Code: Implement to make tests pass (Green phase)
  5. Refactor: Clean up while tests still pass (Refactor phase)
  6. Verify: Run npm run check before committing
  7. Post-mortem: Capture learnings for future tasks (see /perform-post-mortem)