NeuroSploit
An MIT-licensed Rust penetration-testing harness that coordinates model-backed agents, evidence review and reports for authorized security assessments.
What happened lately
NeuroSploit timeline
NeuroSploit is a standalone security-testing harness whose Rust engine coordinates model-backed agents, reconnaissance, evidence checks and reports. Its web console and terminal interface use the same command-line engine. The project describes support for source review, live application assessments and other authorized testing workflows.
The documented installers support Linux, macOS and Windows, install Rust if necessary and build the binary. Users supply model-provider API credentials or an installed and signed-in supported CLI such as Claude Code, Codex, Gemini or Grok. Local model options are also documented. Software licensing is separate from provider charges, subscription limits and the tools required for a particular assessment.
The October 19 version 4.4.0 release describes reconnaissance-aware agent selection, changes to MCP servers during a run, and preservation of vote-rejected findings with an explicit rejected status. These are maintainer-described capabilities; the presence of a report or a model vote does not independently establish a vulnerability.