DEV Community: Matthias Meyer

Claude Mythos Preview found thousands of zero-days in every major OS and browser. Here's what the headlines are missing. published: true

Matthias Meyer — Tue, 07 Apr 2026 22:56:24 +0000

Anthropic dropped something big yesterday. Not a new chat model, not a productivity feature. They revealed that an unreleased model called Claude Mythos Preview has been quietly finding thousands of critical security vulnerabilities across every major operating system and every major web browser. Some of these bugs are 27 years old.

They are not releasing the model publicly. Instead, they launched Project Glasswing, a coalition with Apple, Microsoft, Google, AWS, CrowdStrike, NVIDIA, the Linux Foundation, and others. The goal: patch the bugs before attackers build similar AI.

Most of the coverage I have seen reads like a rewritten press release. The actual details buried in the red team report and system card tell a different story. Here is what stood out to me as someone who runs AI agents in production every day.

The exploit rate is the real headline

Finding bugs is one thing. Every static analysis tool finds bugs. The difference with Mythos is that it does not just find vulnerabilities. It builds working exploits for them.

In testing against Firefox's JavaScript shell, Mythos turned 72.4% of discovered vulnerabilities into successful exploits. It achieved register control in another 11.6% of cases. Previous Claude models could spot bugs but failed almost entirely at exploitation. That gap is gone now.

This is not a scanner. This is a model that reads code, understands the logic, finds the flaw, and writes a proof of concept that works. Autonomously.

Less than 1% is patched

Anthropic says thousands of critical zero-days were found. Fewer than 1% have been fully patched so far. The volume is simply too large for the affected organizations to keep up.

They are publishing cryptographic hashes of vulnerability details today, with plans to reveal specifics after fixes ship. This is standard responsible disclosure, but the scale is unprecedented. We have never had a single tool produce this many verified findings at once.

For context: Google's Project Zero, one of the best human vulnerability research teams in the world, publishes around 50 to 80 bugs per year. Mythos found thousands in weeks.

The CyberGym benchmark gap

On the CyberGym evaluation benchmark, Mythos scored 83.1%. Claude Opus 4.6, the next best model, scored 66.6%. That is not a marginal improvement. That is a generational jump within the same model family.

For anyone tracking AI capabilities over time, this should raise questions about what the next generation looks like. If the gap between Opus and Mythos is this large, what happens when competitors catch up to where Mythos is today?

What they are actually giving away

Buried in the announcement is something I think matters more long term than the vulnerability findings themselves.

Anthropic is offering free Claude Max subscriptions to any verifiable open source maintainer. Not Mythos access, but Opus and Sonnet, which are still capable security tools. They committed $100M in usage credits for Project Glasswing partners and donated $4M to open source security organizations through the Linux Foundation and Apache Software Foundation.

If you maintain critical open source software and you have no security budget, which describes most open source maintainers, you can apply through the Claude for Open Source program.

This is a smart move. The vast majority of critical infrastructure runs on code maintained by small teams or individual volunteers. Giving them access to frontier AI for code review could prevent more bugs than any single audit.

The pricing tells you something

When Mythos eventually becomes available through the API, it will cost $25 per million input tokens and $125 per million output tokens. That is roughly 5x more expensive than Opus 4.6.

Anthropic is not positioning this as a general purpose model. The pricing alone ensures it will only be used for high value tasks where the cost of missing a bug is measured in millions. Security audits, compliance reviews, infrastructure hardening. Not chat, not content generation.

The uncomfortable question

Anthropic built this. They chose to handle it responsibly with controlled access and coordinated disclosure. But the capability exists now. It is a matter of time before other labs, or well funded adversaries, reach similar performance.

The 90 day disclosure window Anthropic set for publishing full vulnerability details is tight. With less than 1% of bugs patched, that creates real pressure on every affected vendor. And the affected vendor list is essentially everyone.

Project Glasswing is a starting point. Anthropic said that explicitly. The real question is whether the defenders can stay ahead when this class of capability becomes widely available.

What this means for developers

If you write code that touches the internet, your threat model just changed. Not because of anything you did wrong, but because the cost of finding vulnerabilities in your code just dropped by orders of magnitude.

The practical takeaways:

Audit your dependencies. If Mythos found 27 year old bugs in OpenBSD, your npm packages are not immune. The bugs that survived decades of human review are exactly the kind AI excels at finding.

Watch for the patches. Over the next 90 days, expect a wave of critical security updates across operating systems, browsers, and open source projects. Apply them quickly.

If you maintain open source, apply for Claude Max. Free access to Opus for security review is genuinely useful. Take it.

Rethink your security testing. Static analysis tools that pattern match against known vulnerability types are not enough anymore. The bar just moved.

The age of AI finding bugs faster than humans can fix them is not coming. It arrived yesterday.

StudioMeyer Crew: Turn Claude into a CEO, CFO, or CTO — Zero API Cost

Matthias Meyer — Tue, 07 Apr 2026 19:50:38 +0000

Switching between "act as my CFO" and "now think like a CTO" prompts works -- sort of. But every time you switch, Claude loses the role. The analysis gets shallow. There's no framework, no structure, no memory of what the CEO said when the CFO takes over.

StudioMeyer Crew is an MCP server that gives Claude 8 expert personas with real business frameworks. Activate a CEO, and Claude uses Porter's Five Forces. Switch to CFO, and Claude runs unit economics with your actual numbers. Zero extra API cost -- runs entirely in your Claude subscription.

What does Crew actually do?

Instead of writing long system prompts yourself, you activate a persona with one command:

"Activate the CEO persona."

Claude instantly adopts the role -- with domain-specific frameworks, output formats, decision patterns, and constraints. Not a generic "act like a CEO" prompt, but a battle-tested system prompt with few-shot examples and anti-patterns.

8 Built-in Personas

Persona	Category	What they bring
CEO	Business	Strategy, vision, Porter's Five Forces, SWOT, delegation frameworks
CFO	Business	Unit economics, ROI analysis, pricing models, budget planning
CMO	Business	Growth strategy, AARRR funnel, content marketing, SEO/GEO
CTO	Tech	Architecture review, DORA metrics, tech debt assessment, code review
PM	Business	RICE/ICE scoring, user stories, PRD generation, JTBD framework
Analyst	Ops	Market research, competitive intelligence, data analysis
Support	Ops	Customer communication, FAQ writing, onboarding guides
Creative	Creative	Brand voice, copy, storytelling, design briefs

10 Tools

Tool	What it does
`crew_activate`	Switch to a persona (loads frameworks + memory context)
`crew_deactivate`	Back to default Claude
`crew_list`	Show all available personas
`crew_status`	Who am I right now? Duration, token estimate
`crew_feedback`	Rate the persona (1-5)
`crew_create`	Build your own custom persona
`crew_delete`	Remove a custom persona
`crew_workflow_list`	Available multi-persona workflows
`crew_workflow_run`	Run a full workflow (e.g. CEO then CFO then CTO)
`crew_guide`	Help and onboarding

Getting started: 30 seconds

Add the server URL in Claude Desktop or Cursor:

https://crew.studiomeyer.io/mcp

Or via Claude Code:

claude mcp add crew --transport streamable-http https://crew.studiomeyer.io/mcp

Works with Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, and Zed.

What you can actually do with it

Strategy session in 5 minutes

"Activate CEO. Review our pricing strategy: we have 4 MCP products ranging from free to $49/month."

Claude as CEO runs a structured analysis: market positioning, competitive landscape, pricing psychology, revenue projections. Not a vague opinion -- a framework-driven assessment with clear recommendations.

Financial sanity check

"Switch to CFO. Are these unit economics sustainable?"

The CFO persona calculates: customer acquisition cost, lifetime value, margin per tier, break-even timeline. With actual numbers, not hypotheticals.

Multi-persona workflows

The real power: chaining personas. Three built-in workflows:

Workflow	Personas	What happens
Strategy Review	CEO, then CFO, then CTO	CEO defines strategy, CFO validates finances, CTO checks feasibility
Content Pipeline	CMO, then Analyst, then Creative	CMO plans content, Analyst fact-checks, Creative polishes
Product Launch	Analyst, PM, CEO, CMO, CTO	Full launch analysis from research to go-live checklist

"Run the strategy review workflow for our Q2 plans."

Each persona builds on the previous one's output. The CFO doesn't repeat what the CEO said -- they challenge it.

Memory Bridge

If you also use StudioMeyer Memory, Crew automatically loads relevant context when activating a persona. The CFO searches for "revenue, costs, pricing decisions" in your memory. The CTO loads "architecture, tech stack, dependencies."

No manual context switching. The persona knows your history.

What makes this different

Feature	StudioMeyer Crew	DollhouseMCP	Ruflo
Personas	8 (deep, with frameworks)	150+ (generic)	60+ (swarm)
Memory integration	StudioMeyer Memory	None	Own DB (AgentDB)
Multi-persona workflows	3 built-in	None	Swarm topology
Custom personas	Yes (Markdown + YAML)	No	No
API cost	$0 (Claude subscription)	$0	Requires API keys
Domain frameworks	DORA, Porter, AARRR, RICE	Generic prompts	Generic prompts

The key difference: Crew personas aren't just role descriptions. They include decision frameworks, few-shot examples, output format constraints, and anti-patterns. A CFO persona that says "consider ROI" is useless. A CFO persona that runs a 5-year DCF model with your inputs is useful.

Custom personas

Don't need a CFO but need a Legal Advisor? Create your own:

"Create a persona called Legal Advisor, focused on contract review, GDPR compliance, and risk assessment."

Crew saves it as a Markdown file with YAML frontmatter. Your custom personas persist across sessions and work exactly like built-in ones.

Pricing

Free. All 10 tools, all 8 personas, all 3 workflows. No API keys, no credit system, no hidden costs. Runs in your existing Claude subscription.

We built Crew because we got tired of writing the same "act as my CFO" prompts and getting shallow results every time. Eight personas, tested across hundreds of real business decisions at StudioMeyer. If you want Claude to think like an expert instead of a generalist -- try it.

Learn more about StudioMeyer Crew