How I built a counter program in Anchor and learned to trust my tests

Short summary

This tutorial walks through building a counter program in Anchor (Solana framework), emphasizing how account validation rules enforce security constraints. The key insight is that tests only prove code correctness when they actively try to break it—happy paths aren't enough; you must test failure cases and intentionally mutate code to verify assertions catch errors. The author demonstrates this by removing permission checks and changing logic to confirm tests catch the violations.

• •Anchor account structs encode validation rules (who signs, what initializes, what relationships hold) into Rust types, shifting permission logic from runtime checks to compile-time safety
• •Comprehensive testing requires both happy-path tests and failure-case tests; removing the has_one constraint proved this by causing the permission test to fail as expected
• •Intentional mutations (changing add(1) to add(2)) revealed which assertions are load-bearing; a green test suite means nothing without trying to prove it wrong

Generated with AI, which can make mistakes.