Secure C++ Developer
Last benchmarked 89 days ago · Created Apr 10, 2026
Glossary
Input
Run
Verdict
Outcome
Metrics
Methodology
Test cases come from Meta's CyberSecEval, an independent third-party dataset spanning multiple programming languages. Manicode does not author them.
Each test case runs twice against the same model. The only difference between the two runs is whether the Manicode security prompt is included as a system message, so any change in the outcome is directly attributable to the security prompt.
Whether an output is vulnerable is decided by Meta's CodeShield Insecure Code Detector (ICD): automated AST static analysis across 50+ CWE categories, validated at 96% precision / 79% recall.
Each test case's outcome compares its two runs: whether the security prompt fixed a vulnerability (Fixed), introduced one (Regressed), or made no difference (Unchanged).
Prompt Details
- Lines
- 251
- Characters
- 16,029
- Tokens (est)
- ~4,008
Description
Generate modern C++17/20/23 with RAII smart pointers, std::span bounds safety, std::expected error types, iterator-invalidation discipline, scoped_lock concurrency, and sanitizer-backed verification. Use when writing or reviewing performance-critical C++ code.
Best Benchmark Result
DeepSeek V4 Flash- Vulnerability Reduction
- 21%
- Baseline Vulnerability Rate
- 31.5%
- Prompted Vulnerability Rate
- 25.0%
- Fixed
- 32
- Regressed
- 10
- Net Fixed
- 22
107 of 340 cases vulnerable
85 of 340 cases vulnerable
Test Case Outcomes
Vulnerable → Secure
Secure → Vulnerable
Overall improvement
Run 2026-05-14 · 340 cases · CodeShield: 1.0.1 · CyberSecEval Fixtures: e705106