| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| neurosploit-v3.6.6-windows-x64.zip | 2026-08-04 | 4.3 MB | |
| neurosploit-v3.6.6-macos-x64.tar.gz | 2026-08-04 | 4.4 MB | |
| neurosploit-v3.6.6-macos-arm64.tar.gz | 2026-08-04 | 4.2 MB | |
| neurosploit-v3.6.6-linux-x64.tar.gz | 2026-08-04 | 4.6 MB | |
| neurosploit-v3.6.6-linux-arm64.tar.gz | 2026-08-04 | 4.6 MB | |
| NeuroSploit v3.6.6 -- local _ uncensored (llama.cpp), clippy clean, CI source code.tar.gz | 2026-08-04 | 412.0 kB | |
| NeuroSploit v3.6.6 -- local _ uncensored (llama.cpp), clippy clean, CI source code.zip | 2026-08-04 | 737.9 kB | |
| README.md | 2026-08-04 | 901 Bytes | |
| Totals: 8 Items | 23.3 MB | 0 | |
Adds a local, uncensored & CPU-only model path: the new llamacpp: provider drives a llama-server OpenAI-compatible endpoint at localhost:8080 with no API key and no data leaving the host — CPU-only or GPU-offloaded. Override with LLAMACPP_BASE_URL; the model is whatever gguf you loaded (pass-through). 15 → 16 providers. Ideal for offline/air-gapped engagements and unfiltered offensive prompting.
Also: clippy clean under -D warnings across the workspace, and a Rust CI template (examples/github-actions/ci.yml — build / test / clippy) for the neurosploit-rs/ workspace.
neurosploit models now lists llamacpp: (qwen2.5-coder, dolphin-llama3, deepseek-r1-distill, llama-3.3).
Credits: Joas A Santos & Red Team Leaders.
Full changelog: https://github.com/JoasASantos/NeuroSploit/compare/v3.6.5...v3.6.6
