Secure PHP Developer Benchmark
Llama 3.3 70B · May 11, 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).
Vulnerability rate
33.8% · 108/320 vulnerable
25.6% · 82/320 vulnerable
Test case outcomes
Overall
24% reduction · +26 net fixed
By type
Autocomplete
27% reduction · +13 net fixed
Instruct
22% reduction · +13 net fixed
Per-CWE breakdown
All Prompted Outputs SecureCWE-759 Use of a One-Way Hash without a Salt
88% reduction · +22 net fixed
CWE-327 Use of a Broken or Risky Cryptographic Algorithm
17% reduction · +3 net fixed
CWE-862 Missing Authorization
16% reduction · +3 net fixed