Skip to content

CLAUDE.md

Latest commit

 

History

History
65 lines (46 loc) · 2.96 KB

CLAUDE.md

File metadata and controls

65 lines (46 loc) · 2.96 KB

Aurora Instructions

These instructions are for AI assistants working in this project.

Always open @/.aurora/AGENTS.md when the request:

  • Mentions planning or proposals (words like plan, create, implement)
  • Introduces new capabilities, breaking changes, or architecture shifts
  • Sounds ambiguous and you need authoritative guidance before coding

Use @/.aurora/AGENTS.md to learn:

  • How to create and work with plans
  • Aurora workflow and conventions
  • Project structure and guidelines

MCP Tools Available

Aurora provides MCP tools for code intelligence (automatically available in Claude):

lsp - LSP code intelligence with 3 actions:

  • deadcode - Find unused symbols, generates CODE_QUALITY_REPORT.md
  • impact - Analyze symbol usage, show callers and risk level
  • check - Quick usage check before editing

mem_search - Search indexed code with LSP enrichment:

  • Returns code snippets with metadata (type, symbol, lines)
  • Enriched with LSP context (used_by, called_by, calling)
  • Includes git info (last_modified, last_author)

When to use:

  • Before edits: Use lsp check to see usage impact
  • Before refactoring: Use lsp deadcode or lsp impact to find all references
  • Code search: Use mem_search instead of grep for semantic results
  • After large changes: Use lsp deadcode to find orphaned code

Keep this managed block so 'aur init --config' can refresh the instructions.

Dev Rules

POC first. Always validate logic with a ~15min proof-of-concept before building. Cover happy path + common edges. POC works → design properly → build with tests. Never ship the POC.

Build incrementally. Break work into small independent modules. One piece at a time, each must work on its own before integrating.

Dependency hierarchy — follow strictly: vanilla language → standard library → external (only when stdlib can't do it in <100 lines). External deps must be maintained, lightweight, and widely adopted. Exception: always use vetted libraries for security-critical code (crypto, auth, sanitization).

Lightweight over complex. Fewer moving parts, fewer deps, less config. Simple > clever. Readable > elegant.

Open-source only. No vendor lock-in. Every line of code must have a purpose — no speculative code, no premature abstractions.

Project Specifics

  • What: Vanilla JS library — ADB-direct Android device control for autonomous agents. Accessibility tree in, pruned snapshot out.
  • Language: Vanilla JavaScript, ES modules, no build step
  • Runtime: Node.js >= 22
  • Protocol: ADB (Android Debug Bridge) via child_process.execFile — no Appium
  • Device: Any ADB-connected Android device or emulator
  • Modules: 8 files in src/, zero required deps
  • Tests: 94 passing — run with node --test test/unit/*.test.js test/integration/*.test.js
  • Docs: docs/README.md (navigation guide to all documentation)

@MEMORY.md