🚀 A curated list of cutting-edge cybersecurity tools showcased at the Black Hat Arsenal events — covering offensive, defensive, and research-focused security utilities.
Whether you're in red teaming, blue teaming, appsec, or OSINT — this list helps you explore and leverage the best tools demonstrated live by security professionals across the world.
- Tools are grouped by the location of the Black Hat event (USA, Europe, Asia, Canada, MEA), then by year.
- Each year has its own page where tools are grouped by track category, with description, speakers and repository link.
- Two cross-event indexes let you browse the whole collection by track or by name.
- 🔎 Search — filter every tool by name, speaker, region, year and track
- 🗂️ All tools by track — the 20 Arsenal tracks, every event
- 🔤 All tools A–Z — one line per tool, with every event it was presented at
- 2015 — 12 tools
- 2016 — 18 tools
- 2017 — 16 tools
- 2018 — 23 tools
- 2019 — 23 tools
- 2020 — 26 tools
- 2021 — 22 tools
- 2022 — 22 tools
- 2023 — 43 tools
- 2024 — 58 tools
- 2025 — 51 tools
- 2026 — 49 tools
- 2014 — 12 tools
- 2015 — 31 tools
- 2016 — 33 tools
- 2017 — 37 tools
- 2018 — 38 tools
- 2019 — 38 tools
- 2020 — 23 tools
- 2021 — 41 tools
- 2022 — 37 tools
- 2023 — 61 tools
- 2024 — 64 tools
- 2025 — 59 tools
- 2013 — 48 tools
- 2014 — 54 tools
- 2015 — 58 tools
- 2016 — 91 tools
- 2017 — 87 tools
- 2018 — 89 tools
- 2019 — 91 tools
- 2020 — 51 tools
- 2021 — 64 tools
- 2022 — 83 tools
- 2023 — 90 tools
- 2024 — 93 tools
- 2025 — 126 tools
- 2026 — 101 tools
The whole list is available as tools.json and tools.csv (2098 rows: name, description, URL, tracks, speakers, year, location). Both are regenerated from the per-tool JSON files under tools/ — treat them as read-only.
We welcome community contributions to make this list better!
- 📁 Each tool is one JSON file at
tools/{LOCATION}/{YEAR}/<tool-name-slug>.jsonwith: Tool Name, Description, GitHub URL (if available), Tracks, Speakers. - 📝 Follow CONTRIBUTING.md for the format and the list of valid track names.
- ✅ Run
python3 scripts/normalize.py,python3 scripts/validate.py(must report 0 errors) andpython3 AutoReadme.py(regenerates every README, index and data file), then open a pull request.
⚠️ All README files, the indexes andtools.json/tools.csvare generated — edit the JSON files, not these.
