SMT Solver Interface
SMT (Satisfiability Modulo Theories) solvers combine SAT solving with decision procedures for theories like arrays, bitvectors, and arithmetic. They are essential tools for program verification and synthesis.
When to Use This Skill
- •Verifying program correctness
- •Solving constraint problems
- •Automated theorem proving
- •Program synthesis
- •Test generation
What This Skill Does
- •Constraint Encoding: Translate program properties to SMT formulas
- •Solver Invocation: Call SMT solvers (Z3, CVC5, etc.)
- •Model Extraction: Get satisfying assignments
- •Unsat Cores: Identify conflicting constraints
- •Incremental Solving: Reuse solver state
Key Concepts
| Concept | Description |
|---|---|
| SAT | Boolean satisfiability |
| SMT | SAT + theories (arrays, bitvectors, etc.) |
| Theory | Decision procedure for specific domain |
| Model | Satisfying assignment |
| Unsat Core | Minimal unsatisfiable subset |
Tips
- •Use incremental solving for efficiency
- •Choose appropriate theories for your domain
- •Use quantifiers sparingly (expensive)
- •Extract unsat cores for debugging
- •Profile solver time for large problems
Common Use Cases
- •Program verification
- •Test generation
- •Program synthesis
- •Configuration analysis
- •Security analysis
Related Skills
- •
hoare-logic-verifier- Uses SMT for verification - •
model-checker- Alternative verification approach - •
symbolic-execution-engine- Generates SMT constraints - •
refinement-type-checker- SMT-based checking
Canonical References
| Reference | Why It Matters |
|---|---|
| De Moura, Bjørner "Z3: An efficient SMT solver" | Z3 paper |
| Barrett, Tinelli "Satisfiability Modulo Theories" | SMT overview |
| Kroening, Strichman "Decision Procedures" | Comprehensive book |
Tradeoffs and Limitations
Approach Tradeoffs
| Approach | Pros | Cons |
|---|---|---|
| Z3 | Powerful, well-documented | Can be slow |
| CVC5 | Good for theories | Less common |
| Mono-solving | Simple | Not incremental |
When NOT to Use This Skill
- •Decidable problems (use specialized algorithms)
- •When model checking is more appropriate
- •Very large formulas (may not scale)
Limitations
- •Quantifiers can cause undecidability
- •Non-linear arithmetic is hard
- •Solver performance varies
Assessment Criteria
A high-quality implementation should have:
| Criterion | What to Look For |
|---|---|
| Correctness | Sound encoding |
| Efficiency | Appropriate theory selection |
| Usability | Clear error messages |
| Robustness | Handles solver timeouts |
Quality Indicators
✅ Good: Correct encoding, incremental solving, model extraction ⚠️ Warning: Incorrect encoding, no timeout handling ❌ Bad: Soundness bugs, crashes on large inputs
Research Tools & Artifacts
Real-world SMT solvers:
| Tool | Why It Matters |
|---|---|
| Z3 | Microsoft SMT solver |
| CVC5 | SMT solver |
| Boolector | Bitvector solver |
| Alt-Ergo | Proof assistant |
Key Solvers
- •Z3: Production SMT
- •CVC5: Research solver
Research Frontiers
Current SMT research:
| Direction | Key Papers | Challenge |
|---|---|---|
| Arithmetic | "Nonlinear SMT" | Reals |
| Quantifiers | "Quantifier Instantiation" | Automation |
Hot Topics
- •SMT for verification: Program verification
- •SMT for synthesis: Program synthesis
Implementation Pitfalls
Common SMT bugs:
| Pitfall | Real Example | Prevention |
|---|---|---|
| Encoding | Wrong encoding | Verify |
| Timeout | Solver timeout | Timeout |