mirror of
https://github.com/CyberSecurityUP/NeuroSploit.git
synced 2026-09-29 12:31:43 +02:00
CyberSecurityUPandClaude Opus 5
9d83cb6e30
Backend ------- - knowledge_graph.rs — the durable structure under attack_graph's per-run view: typed entities (asset/endpoint/weakness/technique/finding/account/credential/ impact) joined by typed, weighted, provenance-carrying edges, accumulated across runs in .neurosploit/graph.json plus a per-run copy the report and web console can draw. Answers what a finding list can't: ranked attack paths, and the frontier of entities observed but never proven — where chaining should look next. Agents only sometimes fill chains_from, so progression is also inferred between adjacent kill-chain stages; those edges are marked inferred, weighted lower, and drawn dashed, because presenting a hypothesis as evidence is the graph lying about itself. Secrets stay in the vault, never the graph. - memory.rs — four tiers scoped by lifetime, not importance: working (one run), engagement (one target), technique (one agent/CWE), reusable (generalized). Promotion is evidence-gated and needs independent evidence at each step: a claim repeated within a run becomes engagement knowledge; one confirmed across runs becomes technique knowledge; one that held on two DIFFERENT targets is generalized into a reusable lesson with host-specific tokens stripped. Nothing is promoted on a single observation, which is exactly what a hallucination looks like. Recall is scored (overlap × past success × recency) and injected into recon/exploit prompts as leads to verify. Recalled memos are credited only when the run they informed actually found something. - rectify.rs — a mistyped command cost a full round trip through /help, at the worst possible moment during a live run. Accepted-as-typed wins over everything (so the /url alias is never "corrected" to /ua), then unique prefix, then Damerau-Levenshtein with a length-scaled budget, and a tie is reported rather than resolved. Arguments too: a bare host gets its scheme, an out-of-range count is clamped with a note instead of silently reverting, a near-miss model id is matched against the live catalog. - pool.rs — when every configured model is exhausted or its token is dead, try whatever else this machine can actually reach (an installed CLI subscription, or a provider whose key is in the environment) before parking. A run that stops on a box with three other usable backends stopped for no reason. - repl.rs — /memory, /forget, /graph; a recovered run resumes by itself where nobody is watching (piped stdin — the web console — or NEUROSPLOIT_AUTO_RESUME), since a `/continue` prompt there waits forever. Web --- - Attack path: the stage list was seven hardcoded values, so findings the harness staged outside it were silently dropped — 5 of 27 on a real run. Rewritten against the harness's own stage list with unknown stages kept, two-line labels (every node used to read "SQL Injection Authent…"), stage column headers, pan/zoom/fit, path highlighting, severity filter, and the run's graph.json used when present. - Dashboard: coverage, findings by severity, top weaknesses, and annualized loss exposure via FAIR — frequency from exploitability × validation confidence, magnitude from assumptions shown on screen and editable, reported as a range. The posture score saturates instead of subtracting, so it keeps discriminating past the first critical. - Run history groups into one folder per target with a filter, instead of one flat list that grows forever. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BvdGy9XtVWSdXDTa3FFLJv
NeuroSploit v4.0.0 — web console
A browser UI for the neurosploit CLI harness: a 5-step engagement wizard (Asset → Scope & Auth
→ Leads → Model & Run → Review), a live structured findings view with a generative attack-path
graph, run history, an Auth & Keys menu, and a real terminal — all driven by spawning the actual
CLI binary, never a reimplementation of harness logic.
- Asset — pick black/white/grey-box, host/infra, or AI/LLM, set the target or repo.
- Scope & Auth — objective, focus, out-of-scope, and a link into the Auth & Keys menu.
- Leads — the categorized agent picker (435 agents auto-classified) + custom leads.
- Model & Run — pick a provider/model from the live catalog, API-key vs. subscription auth mode, votes/chain-depth/recon intensity.
- Review — confirm the plan, then
Start Exploitationspawns the real CLI. - Auth & Keys (one menu, 🔑 in the sidebar) — target auth header + named roles for
IDOR/BOLA/BFLA testing, per-provider API keys (kept in server memory only, never on disk), and
an explicit
creds.yamlpath override. - Generative Attack Path Chaining — findings are grouped into kill-chain columns (recon → initial-access → execution → privesc → lateral → exfil → impact) with chained findings linked back to their parent, built live as findings stream in.
- Terminal dock (xterm.js) —
❭_in the sidebar, the topbar button, orCtrl+\`` opens a docked terminal running a realneurosploitREPL session. Its stdout is streamed **unstripped**, so the harness's own colour and box-drawn panels render as they do in a local shell. Line editing (echo, ←/→, history,Tabslash-command completion,Ctrl+C/L/U/A/E`) is local because the child is spawned over a pipe, not a PTY, and therefore never echoes. The target picker in its header switches between a standalone session and the running engagement, so mid-run instructions go to the same process that is doing the testing. - Real REPL underneath run/whitebox/greybox — the wizard scripts an actual interactive
neurosploitsession instead of a one-shot CLI call, so it keeps reading stdin while the engagement streams. The Activity log tab grows a❭prompt box to send/status,/stop,/continue, or a plain-language instruction mid-run.host/aitest/skillsstay one-shot (their onboarding scope picker can't be scripted over piped stdin).
cd neurosploit-rs && cargo build --release # build the CLI once
node web/server.js # → http://localhost:4173
Zero npm dependencies (Node ≥18, built-ins only: http, child_process, events, fs).
API reference: API.md.
Layout
web/
├── server.js backend: static server + agents_md/runs reader + CLI process manager
├── public/
│ ├── index.html SPA shell
│ ├── style.css lead-board / live-run / terminal-dock styling
│ ├── app.js client logic (fetch + EventSource + terminal, no framework)
│ └── vendor/ xterm.js + fit addon (vendored; nothing is fetched at runtime)
├── API.md
└── package.json