Skip to main content

Changelog

May 2026 — v2.2 (Current)​

Technical accuracy

  • Added exact HexStrike Kali package version to installation version table: 0.0~git20260306.8333779
  • Gemini CLI and Ollama table entries clarified: 0.1.x (latest 0.1 at test time) / 0.3.x (latest 0.3 at test time)
  • Remaining personal lab IPs (172.16.59.128, 172.16.59.129, 172.16.59.132) and username (andrey) fully replaced with <target-ip> / <your_user>@<kali-ip> placeholders across installation and gemini pages
  • Web & Cloud: second HexStrike MCP config block corrected to port 8889 — both blocks now consistent (Burp Suite on 8888, HexStrike on 8889)

Framing

  • AD page subtitle "Fully Automated AD Discovery and Exploitation" → "AI-Assisted AD Discovery and Exploitation … Lab walkthrough on GOAD-Mini"
  • ADCS: remaining ### The One-Prompt Approach section renamed to ### Single-Prompt Initiation; ### Cursor + Hexstrike. Fully Automated… heading corrected to (Lab) variant
  • ADCS: "AI determines next steps" → "AI selects the next tool based on prior output"
  • Password disclaimer title updated from "Passwords Cracking: Full Guide with Real-Life Examples" to "AI-Assisted Password Security Auditing"
  • Password page TOC labels renamed to match current framing (Authorized Audit Methods, Credential Exposure Review, Defensive Hardening Resources, etc.)
  • Homepage subtitle updated: "HexStrike does the rest" → "HexStrike assists with tool selection, execution, and reporting"

May 2026 — v2.1​

Technical accuracy

  • Fixed OpenAI Codex version in Tested Environment table: 0.1.x → 0.77.0 (installation page)
  • Normalized Web & Cloud page ports: both HexStrike MCP configs now consistently use port 8889 (Burp Suite occupies default 8888)
  • Replaced personal IP (172.16.59.132) and SSH key name (hexstrike_kali) with generic placeholders across installation, gemini, and ollama pages

Framing

  • ADCS ESC8 article renamed: "One-Prompt Domain Compromise" → "Single-Prompt ADCS ESC8 Attack Chain"
  • "Fully automated" / "AI makes all strategic decisions" replaced with observed-lab language across AD and ADCS pages
  • Old disclaimer title "Passwords Cracking: Full Guide with Real-Life Examples" updated to match current page title

Coverage

  • :::info Last tested banner added to all 16 technical pages that were missing it
  • Known Limitations section added to 5 lab/walkthrough pages (lab-setup, full-subnet, black-box-ad, active-directory, adcs-esc8)

May 2026 — v2.0​

Safety & framing

  • Added site-wide authorization banner on every page
  • Added per-page authorization disclaimer to all attack technique and walkthrough pages
  • Reframed homepage: "AI-assisted orchestrator" replaces "autonomous operator"
  • Reframed AD and ADCS articles: "lab-observed, approval-gated" replaces "fully autonomous / no human intervention"
  • Full PT guide rewritten as "Controlled Lab Assessment Workflow" with explicit authorization gates at each phase
  • Offensive prompts replaced with structured templates: objective / authorization gate / evidence / detection / remediation

Redaction

  • AD page: lab credentials, NTLM hashes, AS-REP hashes, krbtgt hash redacted
  • SMB page: raw credential list replaced with generalized description
  • All /home/andrey/ paths replaced with generic placeholders
  • SSH key name generalized to ~/.ssh/<your_key>

Content quality

  • Defensive hardening sections added to AD and ADCS pages (event IDs, hardening checklists, tooling)
  • Benchmark methodology appendix added to HackerAI comparison (versions, run count, success criteria, limitations)
  • "Last tested" version banners added to all LLM integration and attack technique pages
  • Password security page retitled and reframed as "Authorized Password Security Auditing"
  • "AI determines optimal next steps" → "proposes next steps for operator review"

Technical fixes

  • Repo URL fixed: HexStrike/hexstrike-ai → 0x4m4/hexstrike-ai
  • Quick-start corrected: hexstrike_server / hexstrike_mcp (matches Kali package)
  • All Google-search placeholder links replaced with direct Medium URLs
  • Password math formulas corrected: 10⁴, 26⁵
  • MCP config argument fixed: - server (Unicode dash) → --server
  • Port 8889 in web+cloud page labeled and explained
  • HexStrike AI capitalization normalized to HexStrike AI
  • "Press enter or click to view image" Medium residue removed (5 instances)
  • All "Promt" typos fixed (7 files)
  • Tested version matrix added to installation page

Infrastructure

  • Link-check CI job added (runs on push to main and weekly)
  • About page added with author bio, social links, donate link
  • Changelog added

December 2025 — v1.0 (Initial)​

Initial publication — complete Docusaurus site built from Medium article exports.