Deterministic code-intelligence — a capability fingerprint per file, no model, no GPU, no network.
Gamma builds a high-dimensional capability fingerprint per file and binds distributed capability signals across files into a single coordinated-pattern verdict — backdoor and exfiltration shapes that no single file reveals. Precision-favoured by design. No model inference, no GPU, no network at scan time.
Numbers from the 2026-05-08 blind test (24 risk fragments in 5 groups + 50 innocent decoys) and the 8-repository real-world cohort. Reproducible from the pinned research corpus.
Zero false alarms on 50 decoy files and 8 public repositories. Per-file recall is intentionally low (0.208): one flagged file is enough to triage the surrounding module. A recall-tuned profile trades precision for coverage — stated, not hidden.
Internal rule names, internal weights, the internal scoring mechanics and formula identifiers stay sealed. Known gaps (availability-class sabotage, distributed-ransomware binding) are disclosed under evaluation, not advertised.
Portal precision 1.0 is blind-test scope — recall 0.208 by design; do not mix in-tree regression KPI corpora.
Normative files, specs, and conformance guarantees from the official repository.
Deterministic code-intelligence. High-dimensional capability fingerprint per file.
High-dimensional capability fingerprint. 1.000 precision, 0 false alarms.
Deterministic re-run guarantee — 3/3 identical across runs.
Proprietary commercial license. Use, copying, or reverse engineering of the sealed engine core is strictly prohibited. Evaluators must sign the bilateral NDA.