Cyberdefense Hackathon
Explore the projects and technologies built at the hackathon.
Showing 1–14 of 14 projects
Trí Nguyễn
Sentinel is an autonomous cyber-response agent that investigates real security incidents, safely tests remediation, takes approved action, and learns from every attack to respond faster next time.
Anonymous builder
Interlock is an agent security operations platform that monitors autonomous AI agents, blocks unauthorized actions, preserves forensic evidence, and coordinates isolated incident investigations before allowing recovery.
SS HC
An AI red-team that clones your app, attacks the clone through a shield proxy, detects the breach with a ClickHouse SQL query in ~110ms, confirms it with Semgrep, applies a blast-radius-aware fix, and locks the pattern into CI so the same bug can never re-enter.
x
Sentient Stream is a security-agent project exploring how AI can help people understand and respond to cybersecurity risks. Its proposed focus is turning security information into clear, actionable guidance, with an emphasis on transparency and human oversight.
Anonymous builder
Training platform for defensive ai
Luigi Canoro
Albert AI secures AI agents that touch email, code and money. It closes the loop: a verified brief before code is written, Semgrep rules while it is written, attacks against every fix, every successful attack turned into a new rule, and live guarding of every agent action.
Jake Martin
Hacker Bot is an autonomous defensive-security agent that finds and verifies real vulnerabilities on the open web — where every action it takes, and every tool call from any agent you connect, must first pass a deterministic governance gate that answers EXECUTE or REFUSE, with a signed receipt.
shrutika joshi
AgentWatch is a SOC for AI agents. It detects an agent hijacked by prompt injection across 20M events in ~1s with ClickHouse, contains it automatically, proves containment held, and guards LLM-written SQL after finding a ClickHouse readonly bypass.
vincent.leguide
ESCAPE ROOM is an autonomous cyberdefense system that detects AI agent security breaches, investigates root causes, executes policy-authorized remediation, verifies recovery with real Docker probes, and prevents recurring threats, with evidence for every decision.
sunil_sabat
Sitting on a gateway, I watch agentic communication for prompt injection, PII, data theft, unauthorized access etc.
Sreenidhi Hayagreevan
LogWhisperer is an AI security analyst for IT and security teams: ask millions of login logs a question in plain English, by voice or text, and get a clear answer, risk rating, attack timeline, and the exact SQL behind it in seconds.
Aj Mouly
A runtime defense for AI agents that detects a prompt-injection hijack, proves which input caused it, and deploys a guardrail verified against all recorded history before it ships.
Download Download
automation agent for security checking
Byron Lee
Antigense Daisy detects unsafe fail-open AI-worker behavior, gates remediation behind review, and preserves Semgrep, Akash and ClickHouse evidence in verifiable FCO/FCG custody, now with a live local public/private judge review stream.