Skip to content

CLAUDE.md

Latest commit

 

History

History
29 lines (22 loc) · 1.35 KB

CLAUDE.md

File metadata and controls

29 lines (22 loc) · 1.35 KB

Testing Patterns

  • Test Philosophy: Concrete tests validate specific paths, fuzz tests provide confidence across ranges
  • Foundry Gotchas:
    • vm.prank and vm.expectRevert apply to the NEXT external call only
    • Use vm.startPrank/stopPrank for multi-call setups to avoid losing the prank context
  • BTT Structure: Tests use modifiers to match Branch-Based Testing tree structure for clarity
  • Time Advancement: Use _mineBlocks() not vm.warp() (updates both timestamp and block number)
  • Invariant Cache: The cache/invariant directory stores failed sequences for regression testing - only clear after fixing bugs

Fork Testing Setup

# Fork tests use the 'optimism' RPC alias from foundry.toml (uses Infura)
# Ensure .common.env has API_KEY_INFURA set
source .common.env && source .optimism.env
forge test --force  # --force required for OpenZeppelin plugin

Note: Fork tests use vm.createSelectFork("optimism", blockNumber) which references the optimism RPC alias in foundry.toml. This automatically uses your Infura API key from .common.env.

Critical System Constraints

Token Supply Limits

  • WCT Max Supply: 1 billion tokens (no holder > 15%)
  • int128 Safety: Max 100M tokens (1e26) for test amounts to prevent overflows
  • Test Amount Guidelines: whale (10M), large (1M), medium (100k), small (10k)