| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| NeuroSploit v3.6.7 -- Chain _ Exploit (CVE pipeline, chaining, --only, whitebox doctrine) source code.tar.gz | 2026-08-04 | 420.3 kB | |
| NeuroSploit v3.6.7 -- Chain _ Exploit (CVE pipeline, chaining, --only, whitebox doctrine) source code.zip | 2026-08-04 | 749.9 kB | |
| README.md | 2026-08-04 | 1.3 kB | |
| Totals: 3 Items | 1.2 MB | 0 | |
Turns known-CVE hunting into exploitation and makes the harness chain deeper — safely.
- CVE exploitation pipeline (4 new agents):
cve_version_fingerprint→cve_research_analyst(NVD/GHSA mapping) →cve_poc_finder(vet/adapt public PoCs) →cve_exploit_scripter(custom exploit when none exists). - PoCs in the report: agents write runnable proofs to the run's
pocs/folder; a new "Reproduction — PoC scripts" report section lists them, so findings replay end-to-end. - Chaining for any primitive: a chaining doctrine +
chain_cve_to_rce_to_pivotrecipe turn any foothold into the next step (upload→RCE, SSRF→cloud creds, IDOR→takeover, CVE→RCE→pivot), reusing looted creds and reasoning about business logic. Strictly non-destructive (no data loss / DB overwrite / DoS). --only <agent>: re-test exactly one vulnerability, skipping recon selection — onrun/whitebox/greybox.- White-box stays white-box: a static-review doctrine (symbolic
file:linereceipts, source→sink taint, manifest version→CVE) blocks hallucinated live/black-box actions.
435 markdown agents (was 430).
Credits: Joas A Santos & Red Team Leaders.
Full changelog: https://github.com/JoasASantos/NeuroSploit/compare/v3.6.6...v3.6.7
