GPUBoost is a local-first development tool. It analyzes local files, creates reviewable suggestions, and may generate local reports or artifacts. Treat any security or data exposure report with care.

Do not paste secrets, tokens, credentials, private keys, raw private data, private source code, raw diffs, model weights, or generated datasets in public issues, pull requests, comments, logs, or screenshots. Public issues should contain only redacted summaries.

For sensitive reports, use GitHub private vulnerability reporting if it is enabled for this repository. If private reporting is not available, contact the maintainers out-of-band before sharing details. Public reports should include only a redacted, minimal summary.

• API keys, tokens, credentials, private keys, and account identifiers
• Private paths, hostnames, usernames, and organization names
• Raw private source code and raw diffs
• Raw/generated datasets and private benchmark outputs
• Model weights, checkpoints, serialized models, and local databases
• Trial stdout/stderr or logs that include private content

Generated artifacts and raw data are not intended to be committed. The repo ignores generated outputs and common artifact formats such as:

• data/gpuboost/generated/
• data/gpuboost/raw/
• *.pt, *.pth, *.ckpt, *.pkl, *.joblib, *.safetensors, *.onnx
• *.db, *.sqlite, *.sqlite3

Model artifacts are local/generated files. Model predictions are advisory-only and must not apply patches, edit files, override deterministic checks, or replace tests, trials, or benchmark evidence.

GPUBoost does not guarantee speedups or production readiness. Benchmark results depend on hardware, drivers, CUDA/PyTorch versions, power mode, thermals, background load, and workload shape. Synthetic demos are useful validation coverage, not universal proof.