Loading story…
Cloudflare has launched a developer preview that lets websites expose WebMCP interfaces through a single dashboard setting, extending Google Chrome’s emerging proposal for structured AI-agent access to web features. WebMCP is designed to let browser-based agents discover and invoke site-defined tools with JSON Schema inputs and outputs and shared page state, avoiding brittle HTML scraping and simulated clicks. Google has been testing the proposed standard in a Chrome origin trial beginning with Chrome 149, while Cloudflare said its preview currently works in Chrome 145 and later.
Cloudflare’s implementation injects a bridge.js module into pages it serves, where tools can be registered statically or discovered dynamically and execution is proxied back to the site on the same origin. The preview includes two initial tool packs: Content Credentials for reading C2PA image provenance metadata and a Site MCP Server pack that exposes a site’s existing MCP server tools to in-browser agents. The rollout comes with security constraints already outlined in Chrome’s proposal, including requirements for origin-isolated documents, a tools Permissions Policy, and allow="tools" for cross-origin iframe use, while both the standard and Cloudflare’s implementation remain subject to change.
Track how attackers are adapting to this technology.
2 events from the most recent confirmed update back to the earliest known activity.
Cloudflare announced a developer preview that lets websites on its platform expose a WebMCP interface through a dashboard switch. The preview initially included Content Credentials and Site MCP Server tool packs, with support described as limited to Chrome 145 and later.
Chrome made the proposed WebMCP standard available through an origin trial starting with Chrome 149, while also allowing local testing via a Chrome flag for development. The announcement described WebMCP as a browser-mediated way for websites to expose structured tools to AI agents.
Vulnerabilities, threat actors, malware, products, organizations, and breaches Mallory has linked to this story.
Follow how adversaries are adapting to this technology, and where it touches your stack today.
3 references tracked. Mallory keeps watching after this page renders.
infoq.com
Open sourceblog.cloudflare.com
Open sourcedeveloper.chrome.com
Open sourceMap indicators from this story to your assets and identify affected systems in minutes.
Every observed campaign, victim, and pivot linked to actors named in this story.
Malware, exploits, and IOCs connected to the activity described here.
YARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Get matching new stories delivered to your team as they break — not the next morning.
Ask questions about this story and take action on the answers.
Other stories Mallory is tracking that overlap on entities, sources, or topic labels.
Cloudflare has introduced **WriteGuard**, a private beta security layer for Model Context Protocol (**MCP**) servers that intercepts agent requests, enforces tool-specific policies, and logs blocked or failed actions for centralized auditing. The service is designed to curb the risk of AI agents using existing OAuth-backed access to perform sensitive write operations across databases, internal APIs, GitHub, GitLab, Jira, Google Workspace, SaaS platforms, and internal wikis. Cloudflare said WriteGuard can apply consistent controls across multiple MCP servers without requiring changes to each server, while preserving attribution by tying MCP client and session context to the human user and scrubbing sensitive values before audit events are stored. The launch comes as MCP-connected tools are gaining direct write access to production workflows. ElevenLabs recently introduced a hosted MCP connector that lets Claude inspect and modify **ElevenAgents** configurations, including changing prompts, voices, languages, retrieving transcripts, generating sample speech, estimating costs, and even deleting agents from a chat interface. ElevenLabs said it uses OAuth plus organization- and session-level controls and urged users to review tool calls before approval, warning that approval alone does not guarantee safe or correct downstream behavior; the company also pointed teams to testing, versioning, and CLI/API-based rollout practices for stronger governance.
Aug 20, 2026
The **Model Context Protocol (MCP)** has been updated to a fully stateless design in the `2026-07-28` specification, removing required sessions, the `initialize` handshake, and the `Mcp-Session-Id` header so MCP servers can run as standard HTTP services without sticky-session infrastructure. The update also replaces stream-dependent elicitation with **Multi Round-Trip Requests**, adds MCP-specific HTTP headers for gateway and security controls, and tightens authorization through issuer identification and canonical resource audience handling. Cloudflare said its Agents SDK and Workers OAuth Provider already support the new model, and cited production adoption by **Sentry** and **Linear**. At the same time, major vendors including **OpenAI, AWS, Cursor, GitHub, Microsoft, and Vercel** backed **Agent Plugins 1.0.0**, a vendor-neutral format for packaging MCP servers and agent skills for use across clients and runtimes. The push for interoperability comes with growing security concerns: researchers published new **Nuclei** templates to detect publicly exposed AI agent configuration files such as `CLAUDE.md`, `AGENTS.md`, Cursor rules, and `vscode-mcp.json`, warning that misconfigured web roots can leak internal workflows, tool settings, and project conventions as organizations expand AI agent deployments.
Aug 8, 2026
The Model Context Protocol (MCP) has seen significant adoption among engineering teams for enabling large language models (LLMs) and AI coding assistants to interact with services, query documentation, and enhance developer productivity. MCP servers are being used internally across various industries, including regulated sectors like aerospace, to provide non-developers and business stakeholders with controlled access to technical resources. Security remains a critical concern, with best practices emerging for safe deployment, and new tools and frameworks such as FastMCP for Python are gaining traction among developers. AWS has introduced IAM Policy Autopilot, an open-source static analysis tool that functions both as a command-line utility and an MCP server. This tool is designed to help developers and AI coding assistants generate and refine AWS IAM policies by analyzing application code locally, streamlining the process of managing permissions and reducing the time spent troubleshooting access issues. The integration of IAM Policy Autopilot with MCP servers highlights the growing ecosystem of tools leveraging MCP to bridge the gap between AI-driven development and secure, efficient cloud operations.
Jun 29, 2026