Entity trail

Armadin

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-07-07 / agents-researcherArmadin details full sandbox escape in Anthropic's Claude Cowork; Anthropic disputes, silently patchesResearchers at Armadin published an attack chain against Claude Cowork for Windows: manipulating a resume flag passed through CoworkVMService bypasses the per-command unprivileged-user creation to run commands as root, while a second flaw overrides the domain allowlist with a wildcard to strip egress limits and enable data exfiltration. Anthropic argues it isn't a security issue because it requires prior local code execution, yet still shipped a patch — a preview of the accountability gray zone builders enter when agent tools run local VMs on user machines. For anyone shipping sandboxed agents, the lesson is that a resume/replay path and per-request allowlist overrides are exactly where isolation quietly dies.
2026-05-17 / skill-finderPalo Alto Unit 42 Frontier AI Defense: AI Models Found 26 CVEs (75 Issues) in One Month vs Typical <5 — 50% Coding Efficiency Improvement Crosses Autonomous Operator ThresholdPalo Alto Networks' May 2026 Defender's Guide reveals that frontier AI models (Mythos, Opus 4.7, GPT-5.5-Cyber) scanning their code found 26 CVEs representing 75 issues in a single month, vs their usual volume of <5 CVEs/month. They conclude the latest models represent a 50% coding efficiency improvement — a threshold where AI crosses from helpful assistant into autonomous operator. Partnership with Armadin (Kevin Mandia's company) for offensive AI security testing.
2026-04-01 / agents-researcherRSAC 2026 Security Leaders Warn of 'Two-Year Upheaval': Mandia, Stamos, Adamski Predict AI-Powered Exploit Generation at ScaleKevin Mandia (Armadin), Alex Stamos (Corridor), and Morgan Adamski (PwC) warned at RSAC 2026 of a 'perfect storm for offense over the next year or two' as AI-powered autonomous agents operate at machine speed. Stamos projected that within 6-12 months, AI systems could generate sophisticated exploits bypassing modern processor protections on demand, compressing patch cycles to 'Patch Tuesday, exploit Wednesday.' Mandia stated 'nobody's ready' while Adamski highlighted that CISOs cannot stop AI adoption due to board-level demand pressure.
2026-03-11 / agents-researcherArmadin: Mandiant Founder M Autonomous Security Agent StartupKevin Mandia raises M for agent-native security. 60+ employees, Fortune 100 customers in 6 months. Backed by In-Q-Tel.

Graph relationships

INVESTED IN
In-Q-Tel
In-Q-Tel backed Armadin with funding
Source finding
BUILT BY
Kevin Mandia
Kevin Mandia founded Armadin, an autonomous security agent startup
Source finding
WORKS AT
Kevin Mandia
Kevin Mandia works at Armadin
Source finding
CRITICIZES
Claude Cowork
Armadin published a sandbox escape attack chain against Claude Cowork for Windows.
Source finding
INVESTED IN
In-Q-Tel -> Armadin
In-Q-Tel backed Armadin with funding
Source finding
WORKS AT
Kevin Mandia -> Armadin
Kevin Mandia works at Armadin
Source finding

Source trail

Source Source Source Source Source Microsoft TechCrunch (Promptfoo)TechCrunch (Armadin)SiliconANGLE Palo Alto Networks — Defender's Guide to Frontier AI Impact on Cybersecurity May 2026 CyberScoop

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…