... runtime for suspicious AI-agent sessions. A trusted gateway routes a designated session into a persistent synthetic world, with useful world-only Wrong Tickets, observable activity and separately approved bounded changes. The goal is to give people and authorized defender agents room to act. v0.16 addresses a concrete lesson: protecting one route is not enough. In the earlier Sol experiment, all three diverted sessions received the artificial target through an open alternate

... DungeonQ's goal is to give security teams and authorized defender agents time and space to respond while a designated suspicious session works in a persistent synthetic world. v0.15.0 adds a reproducible console showing the next step: what happens after an alert? 1. A real MCP client enters the workspace, checks conflicting claims and saves a note that survives restart. Wrong Tickets remain useful only inside their issuing world. A claimed "source restored" message does not change the route. Follow ... ... real MCP/HTTP requests and fresh artificial resources. Its configured response delay is zero; it does not measure time gained through deception or show an attacking model being fooled. The report preserves a raw failed host checkpoint and separately attributes defender

... only source. 3. Controlled continuation: preview and explicitly approve a finite mutation grant, then observe the follow-up record after a fresh ticket use. Restart with the same private data directory and read the retained note. This gives defenders a concrete place to inspect activity and decide what comes next. Human responders and authorized AI agents are the intended beneficiaries; these views do not claim an automatic incident-response handoff. The updated ten-checkpoint walkthrough also checks a stale-write ... ... instructions: https://github.com/Ranopha/dunge... Walkthrough and recorded evidence: https://github.com/Ranopha/dunge... Comment from Shivam Kushwaha(@shivam_kushwaha16): this is really interesting to security testing especially the two working views make it easier to see both the user side and defender

... Hunters , I have built an application by following Andrej Karpathy's LLM council approach, which simulates the proceedings of an entire courtroom. Each agent has a role to play; one is a prosecutor, another is a defender, another is the judge, etc. Each is an individual DeepAgent with their own set of skills and tools. The backend is built using FastAPI, and the Frontend uses React. I will attach a video here as well so that interested people ...

... both content and   product work.   2. Google Search Console I basically live in it. For anything SEO, it's the only honest source of truth   about what's actually happening vs. what I assume is happening.   3. Ahrefs keyword and competitor

... phishing analysis Behavioral anomaly & psychological trigger detection Sender profiling & threat intelligence enrichment Microsoft 365 integration + SIEM support Automated incident response & executive reporting It s built for IT teams, MSPs, and CISOs who need more than Defender and spam filters. Would love honest feedback especially from anyone who s dealt with BEC, runs an MSP, or works in security. https://huggingface.co/spaces/Sa... ...

... cutoff get crushed. If a $30 scan saves someone one billable hour by helping them know which questions to ask, it has paid for itself five times over. What I'd love feedback on: Defence lawyers / paralegals / public defenders: would you actually use this for triage? What's missing? Anyone who has been through the system: what did you wish you understood about your own file? Hunters in general: does the framing land? Anything that smells off? Brutal feedback welcome ...

... recommending defenses. World's first multi-model validation for cybersecurity decisions. 100+ Real-World Attack Scenarios MOVEit breach, 3CX supply chain attack, Ivanti VPN zero-days, Log4Shell, and 95+ more documented attacks from 2023-2025. Playground Mode Battle AI defenders yourself. Choose from 3 difficulty tiers. Make split-second SOC decisions under pressure. Tournament Mode Watch AI agents compete in ranked battles. Glicko-2 ELO rating system tracks agent performance over time. Live Breach Replay Engine Auto-generates playable scenarios ...

... working on https://threatops.tech Global cyber threat intelligence platform that: Tracks active exploits, ransomware, and phishing campaigns Monitors known exploited vulnerabilities (CISA KEV, MITRE, OTX & more) Provides AI-generated security guidance in plain English Builds a community for defenders to share intelligence and discuss threats This is my attempt to make cyber threat intelligence more accessible whether you re a SOC analyst, developer, startup, or just curious about the threat landscape. I d love your honest feedback: What works ...