Skip to content
View Juwon1405's full-sized avatar
:octocat:
:octocat:

Block or report Juwon1405

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Juwon1405/README.md

Juwon Bang  (優心)

DFIR & Detection Engineering · Tokyo, Japan

email


Building autonomous detection systems and architectural security guarantees. Currently exploring agentic DFIR — MCP-based forensic agents that encode the reasoning pattern of a senior analyst as architecture, not as a prompt.

🔍 Focus

  • Digital Forensics & Incident Response  ·  Windows / macOS / Linux
  • Detection Engineering  ·  MITRE ATT&CK coverage modeling, Sigma
  • DevSecOps & Security Automation
  • Agentic AI for Security  ·  MCP, audit-chained reasoning loops

🛠️ Stack

📌 Featured Projects

🎯 Agentic-DART  flagship — SANS FIND EVIL! 2026

Agentic-DART

Autonomous DFIR agent that thinks like a senior analyst. Architecture-first, not prompt-first. The agent is given a typed, read-only MCP surface — native pure-Python forensic functions plus SIFT Workstation tool adapters (Volatility 3, MFTECmd, EvtxECmd, PECmd, RECmd, AmcacheParser, YARA, Plaso) — spanning broad MITRE ATT&CK enterprise coverage. The full test suite passes on a fresh clone (CI green on Python 3.10/3.11/3.12/3.13). The senior-analyst playbook synthesizes Mandiant + Bianco + Diamond + Palantir ADS + MaGMa UCF + TaHiTI hunt cycle, with every framework block citing its source. Read-only MCP boundary makes destructive ops impossible by construction. Three evaluation tiers: synthetic reference (recall=1.0), noise-injected realistic at ~1:30 IOC:benign (recall=1.0), and NIST CFReDS Hacking Case (community-trusted external benchmark) — v0.6.0 scores 0.50/0.80 strict/lenient on 10 sampled NIST findings, up from v0.5.3's 0.10/0.40 (supply-chain sweeps + collector adapter added in v0.6.0), after parse_registry_hive (issue #52) shipped. Remaining CFReDS gaps (#53/#54/#55) are explicit Phase 2 deliverables. Starts as agentic DFIR; designed to expand toward agentic SOC and beyond.

github.com/Juwon1405/agentic-dart  ·  Submission to SANS FIND EVIL! 2026  ·  MIT

Other projects

🔌 agentic-dart-collector-adapter  new — Phase 1.3

stars last commit python MIT

Velociraptor → evidence_root adapter — stdlib-only Python layer that converts Velociraptor offline-collector ZIPs into the evidence_root layout Agentic-DART consumes. Seeds chain-of-custody (manifest.json + SHA-256 index). Full test suite passes on CI Linux+macOS × py3.10/11/12.

📓 GitNote

stars last commit markdown

GitNote — curated personal knowledge base in InfoSec & computer science. A long-running collection of notes, references, and code snippets from years of DFIR / detection engineering work.

📦 Archived projects

🧪 yushin-gendfir-rag  archived

Unofficial Python replication of Loumachi, Ghanem & Ferrag (2024). RAG + LLM pipeline for DFIR cyber-incident timeline analysis. The work in this repository served as a foundation that informed the design of agentic-dart, which supersedes it with agentic (rather than pure RAG) reasoning and a hardened MCP surface. Kept public as a reference artifact.

Single-file bash DFIR artifact collector for macOS hosts where Velociraptor is not an option. Originator of the supply-chain IOC sweep patterns (litellm PyPI 2026-03, npm typosquat detection) now ported and generalized into agentic-dart. Kept public as a supply-chain reference.

Flask-based macOS DFIR web platform that ingested collector ZIPs & disk images (DD/RAW/E01/AFF/DMG) and produced searchable evidence + PDF incident reports. Paused for post-SANS repositioning as the agentic-dart web UI — reading findings.json + audit.jsonl from an Agentic-DART run and rendering them in the browser.

📖 Published Work

  • Network Attack Packet Analysis for Security Practitioners  ·  보안 실무자를 위한 네트워크 공격 패킷 분석  (co-author, lead)
    Freelec, 2019.11  ·  ISBN 9788965402589  ·  ~370 pp.
    A practitioner's reference covering DDoS, web exploitation, malicious traffic, wireless intrusion, system exploitation, and large-volume packet analysis.
    Yes24  ·  Aladin  ·  Kyobo  ·  Google Books

🏆 Selected Recognition

  • 🥇 Gold Prize, 2017 Korea Open-Source Software Developer Contest  (NIPA, national OSS award)
  • 📜 Patent (filed): Security Event Correlation Analysis Apparatus  (2018, Netmarble Corp.)
  • 🎯 4th place, 2017 CCE National Cyber Defense Competition  (National Intelligence Service of Korea)
  • 🐛 Special Prize, 2015 LINE Bug Bounty Program  (LINE Corp.)

🎥 Community

  • YouTube: DoubleS1405 — long-running Korean-language information-security lecture channel (2014–present)

📚 Curated lists

  • Awesome Stars (GitNote) ⭐ — Starred repos categorized into curated buckets (DFIR / Blue Team / AI / Red Team / Malware / OSINT / etc.), regenerated periodically after curation passes.
  • DFIR — Digital Forensics & Incident Response
  • BlueTeam — Defensive operations & SOC
  • Tools & Tips — Analysis utilities
  • DevSecOps — Security automation & AI
  • Gist — Code snippets

🤝 Open to

Research collaboration · CTF · CSIRT exchange · Open-source security tooling


followers   stars   agentic-dart stars   agentic-dart CI

Pinned Loading

  1. agentic-dart agentic-dart Public

    Agentic-DART — autonomous detection & response agent. Architecture-first, not prompt-first. Starts as agentic DFIR; designed to expand toward agentic SOC and beyond.

    Python 4 4

  2. agentic-dart-collector-adapter agentic-dart-collector-adapter Public

    Thin Python layer that converts Velociraptor IR collector output into the evidence_root layout expected by Agentic-DART.

    Python 1

  3. yushin-mac-artifact-collector yushin-mac-artifact-collector Public archive

    macOS DFIR Artifact Collector — single-file, zero-dependency, modular collection script with selective module execution and supply-chain IOC sweeps.

    Shell 1

  4. yushin-mac-forensics-platform yushin-mac-forensics-platform Public archive

    macOS DFIR Forensics Platform — Flask-based web platform that ingests collector ZIPs and disk images (DD/RAW/E01/AFF/DMG), parses 30+ artifact categories, and produces searchable evidence + PDF inc…

    Python 1