Topic hub
Security
A New Runtime topic hub collecting signals, patterns, field notes, and public sources about security.
Retrieval answer
A New Runtime topic hub collecting signals, patterns, field notes, and public sources about security. Security is tracked here as an evidence-linked topic, not as a static glossary entry. The page connects raw observations to pattern hypotheses, longer analysis, and public sources. Use it as the canonical landing page before drilling into individual records.
Field notes
What should readers understand next?
Stateless MCP Makes Small, Auditable Agent Tools Practical Again
Simon Willison's mcp-explorer, datasette-mcp, and llm-mcp-client show how the stateless specification lowers implementation cost and narrows agent capabilities.
Gemini Spark Moves Browser Agents Into Chrome Sessions
Gemini Spark now integrates with Chrome auto browse, using logged-in browser context with permission while keeping users in the loop for sensitive actions.
LangSmith LLM Gateway Puts Runtime Controls Between Agents and Models
LangSmith LLM Gateway turns spend caps, rate limits, fallbacks, redaction, and provider routing into one governed layer for production agents.
Vercel Passport Makes Identity a Deployment Boundary
Vercel Passport is now generally available, adding verified identity tokens, group claims, audit events, and automation bypasses to protected deployments.
Vercel Sandbox Adds Unix Boundaries for Multi-Agent Work
Vercel Sandbox now supports multiple Linux users and groups, giving each agent a private home directory plus an explicit shared workspace.
Claude Mythos Moves Cryptanalysis Into the Verification Bottleneck
Anthropic's cryptography research shows a frontier model finding HAWK and reduced-round AES attacks quickly, while human validation and disclosure become the scarce production step.
Codex Security CLI Turns Security Review Into a Scannable Workbench
OpenAI's Codex Security CLI packages repository, diff, working-tree, export, validation, and patch flows into a security-review workbench rather than a single scanner command.
DeepsecBench Makes Security Agents a Cost/Recall Tradeoff
Vercel's DeepsecBench reframes security-agent evaluation around recall, precision, cost, total scan time, and a hidden benchmark that resists training leakage.
Cloudflare pvcli Makes Privacy Protocols Debuggable Like Infrastructure
Cloudflare open-sourced pvcli, a curl-like debugger for OHTTP-style privacy flows where no one party is supposed to see the whole request path.
Capital One's VulnHunter Points AI at Real Attack Paths
Capital One's VulnHunter shows a useful shift in AI security tooling: an agent should connect a fix to a reproducible attack path, not only generate a diff.
Raw signals
What changed recently?
Cloudflare: Adversaries aren't breaking in—they’re logging in.
A public X post from Cloudflare with a linked primary source flags Adversaries aren't breaking in—they’re logging in. How is your financial leadership redefining security? Register for our webinar on security resilience.
Anthropic: In a review of our cybersecurity evaluations, we found three incidents in which a Claude model...
A public X post from Anthropic as a public source in its own right flags In a review of our cybersecurity evaluations, we found three incidents in which a Claude model reached the internet from within or while interacting with a third-party evaluation...
Cloudflare: Final breakout block starting now at #CloudflareConnect Sydney!
A public X post from Cloudflare as a public source in its own right flags Final breakout block starting now at #CloudflareConnect Sydney! Heading into deep dives on volumetric attack mitigation, machine-to-machine security, Workers AI, retail showcases,...
OpenClaw: OpenClaw is maturing.
A public X post from OpenClaw with a linked primary source flags OpenClaw is maturing. Today we’re introducing monthly extended-stable releases with backported security and reliability fixes, along with a public maturity scorecard for tracking...
OpenAI: We quietly released the open-source Codex Security CLI, but Hacker News found it before we had...
A public X post from OpenAI with a linked primary source flags We quietly released the open-source Codex Security CLI, but Hacker News found it before we had a chance to share it here... You can now use it to scan repositories, track findings...
Factory: We're firm believers in the open security ecosystem, contributing across our platform: • Publi...
A public X post from Factory with a linked primary source flags We're firm believers in the open security ecosystem, contributing across our platform: • Publishing two open-weight models behind Droid Shield 2.0 • Using Autonomous Security Revi...
GitHub: 5.
A public X post from GitHub with a linked primary source flags 5. Turn on code scanning Code scanning uses CodeQL to identify patterns that lead to vulnerabilities, including injection flaws, unsafe deserialization, and insecure GitHub Action...
Cloudflare pvcli Makes Privacy Protocols Debuggable Like Infrastructure
Cloudflare open-sourced pvcli, a curl-like debugger for OHTTP-style privacy flows where no one party is supposed to see the whole request path.
Cloudflare: Fujitec migrated 3,300+ global employees from legacy VPNs to Cloudflare One (ZTNA) in just 45...
A public X post from Cloudflare with a linked primary source flags Fujitec migrated 3,300+ global employees from legacy VPNs to Cloudflare One (ZTNA) in just 45 days—enhancing security and performance worldwide.
Open Secure AI Alliance Turns the AI-Safety Fight Into a Stack Question
NVIDIA's Open Secure AI Alliance reframes open models, harnesses, identity, safe formats, scanners, and disclosure as shared defensive infrastructure for AI agents.
OpenClaw: NVIDIA has been an incredible partner in helping make OpenClaw more secure.
A public X post from OpenClaw as a public source in its own right flags NVIDIA has been an incredible partner in helping make OpenClaw more secure. Security is a team sport, and open collaboration is how we protect the next generation of agents. Proud...
Vercel Developers: Nuxt 4.5.1 and 3.21.10 fix multiple security vulnerabilities.
A public X post from Vercel Developers with a linked primary source flags Nuxt 4.5.1 and 3.21.10 fix multiple security vulnerabilities. All @nuxt_js users should upgrade as soon as possible. See details, upgrade guidance and our Vercel Firewall response...
Vercel: DeepsecBench evaluates model accuracy, cost, and speed in finding cybersecurity vulnerabilitie...
A public X post from Vercel with a linked primary source flags DeepsecBench evaluates model accuracy, cost, and speed in finding cybersecurity vulnerabilities. Latest results: ▪️ GPT-5.6 Sol scores highest ▪️ Kimi K3 gets half the top score a...
Claude: Opus 5 is available today on all paid plans and the Claude API, priced the same as Opus 4.8.
A public X post from Claude with a linked primary source flags Opus 5 is available today on all paid plans and the Claude API, priced the same as Opus 4.8. It’s the default model on Claude Max, and the strongest on Claude Pro. It’s also offer...
GitHub: Protect your secrets.
A public X post from GitHub as a public source in its own right flags Protect your secrets. 🔐 Join us July 28 for a GitHub Security webinar exploring secret leak trends, public exposure monitoring, and practical steps your org can take to reduce ris...
Google DeepMind: To ensure this model is deployed responsibly, we’re starting with a limited-access pilot for g...
A public X post from Google DeepMind with a linked primary source flags To ensure this model is deployed responsibly, we’re starting with a limited-access pilot for governments and trusted partners. →
LangChain: Agent governance extends beyond security, bringing together authentication, audit logs, user m...
A public X post from LangChain as a public source in its own right flags Agent governance extends beyond security, bringing together authentication, audit logs, user management, provider secrets, data separation, data residency, rate limits, fallbacks,...
Claude: The Claude Security plugin for Claude Code is now available in beta.
A public X post from Claude with a linked primary source flags The Claude Security plugin for Claude Code is now available in beta. Scan your changes for vulnerabilities before you commit, or run a full scan across your codebase, all from you...
OpenAI: We're partnering with @huggingface to investigate an unprecedented security incident.
A public X post from OpenAI as a public source in its own right flags We're partnering with @huggingface to investigate an unprecedented security incident. Cyber-capable OpenAI models compromised Hugging Face production during a benchmark evaluation...
CubeSandbox packages agent execution into isolated microVMs
CubeSandbox provides open-source KVM microVM isolation, lifecycle controls, and per-agent Linux environments.
Cloudflare Gives Agents Temporary Accounts
Disposable identity and scoped access reduce the risk of letting agents touch systems that were designed around long-lived human credentials.
Anthropic / Claude Code Auto Mode: Agent Security Runtime Boundaries
The archive captures Anthropic / Claude Code Auto Mode as a dated public record from Anthropic / Claude Code Auto Mode. It documents agent security expanding from prompt policy into memory, tools, sandboxes, and execution boundaries and is retained as branch-opening evidence for the agent security runtime boundaries trend.
Aitmpl / Dangerous Command Blocker: Agent Security Runtime Boundaries
The archive captures Aitmpl / Dangerous Command Blocker as a dated public record from Aitmpl / Dangerous Command Blocker. It documents agent security expanding from prompt policy into memory, tools, sandboxes, and execution boundaries and is retained as branch-opening evidence for the agent security runtime boundaries trend.
Blog / Securing Agents In Production Agentic Runtime 5191a0715240: Agent Security Runtime Boundaries
The archive captures Blog / Securing Agents In Production Agentic Runtime 5191a0715240 as a dated public record from Blog / Securing Agents In Production Agentic Runtime 5191a0715240. It documents agent security expanding from prompt policy into memory, tools, sandboxes, and execution boundaries and is retained as branch-opening evidence for the agent security runtime boundaries trend.