The maintainers of the official Model Context Protocol Python SDK disclosed a flaw that lets a malicious MCP server trick an application built on the SDK into handing over the OAuth credentials it uses to log in to a real service. Affected versions sent the client secret, authorization code, and PKCE proof key to an attacker-controlled token endpoint, because the SDK did not always verify where the authorization server was. Cycode, which reported it, exchanged the stolen material for a valid access token carrying the app's permissions, and noted the long-lived client secret keeps working until rotated. Fixes are in versions 1.30.0 and 2.2.0.
JFrog disclosed a critical flaw in Bifrost, an open-source AI gateway routing to over twenty LLM providers, that lets an unauthenticated attacker run arbitrary commands on the gateway with a single HTTP request. Tracked as CVE-2026-90898 and rated 9.8, it affects all Bifrost HTTP transport versions before 2.1.0 when management authentication is disabled, which is the default. An attacker registers a stdio-type MCP client through an unauthenticated POST to /api/mcp/client, and Bifrost runs the command immediately, before any handshake, as the gateway user. Because the gateway stores API keys for every connected provider, command execution also exposes those credentials, and the official Docker image binds its management API to all interfaces.
Researchers showed that Amazon Kiro, an AI-powered agentic development environment, can be turned against its user through prompt injection. Attacker-controlled content in a repository the developer opens can steer the Kiro agent into transmitting sensitive local information to an external server, abusing a feature called Kiro Powers that bundles model context protocol server configurations, steering files, and hooks. The developer only has to open the workspace and interact with the agent. It is part of a wider run of similar flaws in AI coding tools, where untrusted content or links quietly redirect an agent into exfiltrating data or executing code without any approval prompt. Updating the tool addresses the reported issue.
Noma Labs found a critical flaw in Ruflo, an open-source platform for orchestrating swarms of AI agents that acts as a harness for coding agents like Claude Code and Codex. Tracked as CVE-2026-59726 and scored 10.0, the flaw is that Ruflo's default Docker deployment exposed its Model Context Protocol bridge to the network with no authentication, offering 233 callable tools. A single unauthenticated request could invoke a shell command tool to run code in the bridge container, read AI provider API keys, steal stored conversations, and poison the platform's persistent agent memory, which can steer future outputs even after the attacker leaves. The flaw is in Ruflo, not in the coding agents.
Offensive security firm Manifold Security detailed a confused deputy flaw in Microsoft's official Azure DevOps MCP server, which lets AI agents read and operate Azure DevOps on a user's behalf across pull requests, pipelines, wikis, and work items, using that user's own permissions. One of the server's tools returns pull request descriptions without the prompt injection guardrail Microsoft had already applied to other tools. Because descriptions accept Markdown, an attacker can bury instructions in an HTML comment: the web interface renders it as nothing, so a reviewer sees an ordinary change, while the REST API returns the text verbatim and hands it straight to the agent.
Researchers at Intezer showed that a poisoned web page could take over AWS Kiro, an AI coding tool, by getting the agent to rewrite the file that governs what it is allowed to run. A prompt injection in fetched content drops custom code into the MCP settings file, which executes the moment the file is saved. The same write to execution path was demonstrated on Kiro's release day in July 2025 by another researcher, who also flagged writing to a Visual Studio Code settings file to allowlist shell commands. AWS added an approval prompt, but only in supervised mode, leaving the default autonomous mode writing the file unprompted.
Researchers at Island uncovered FakeGit, a campaign running roughly 7,600 malicious GitHub repositories from about 6,600 lookalike developer profiles, of which more than 800 pose as AI skills or Model Context Protocol servers. The fake projects borrow the names and workflows of familiar tools, covering Gmail and WhatsApp integrations through to Databricks, Jenkins, and Docker tooling, and their convincing README files walk a user or an agent from routine setup into downloading a malicious ZIP. That triggers a loader chain ending in SmartLoader, which establishes persistence and pulls further payloads such as the StealC infostealer.
Microsoft is warning that attackers can hijack AI agents through poisoned tool descriptions, the plain-text notes that tell an agent what a tool does. Because agents connect to systems through the Model Context Protocol and read these descriptions to decide how to act, an attacker who updates a trusted third-party tool can bury a hidden instruction in its description, telling the agent to quietly collect and exfiltrate data on its next task. Many setups pick up description changes without re-approval, so the poisoned version goes live silently. Each step the agent takes looks legitimate and runs with the user's own permissions, so no alarm fires.
Wiz Research found a high-severity flaw in Amazon Q Developer, Amazon's AI coding assistant, that let a malicious code repository run commands and steal a developer's cloud credentials simply by being opened. The bug (CVE-2026-12957) lay in how Amazon Q handled Model Context Protocol servers: it read an MCP configuration file from the open workspace and automatically launched the servers it defined. Because those servers run as local processes that inherit the developer's full environment, a single config file committed to a repo could reach AWS keys, cloud tokens, API secrets, and SSH agent sockets, turning a git clone into a full compromise. Amazon has patched the issue and published an advisory.
Researchers at Tenet Security have disclosed Agentjacking, a new attack that turns AI coding assistants like Claude Code, Cursor, and Codex into tools for running an attacker's code on a developer's machine. The trick abuses Sentry, a widely used error-tracking service: anyone can submit a fake error event using a project's DSN, a public write-only key embedded in website code, and the AI agent, fetching that event through Sentry's MCP integration, cannot tell the malicious instructions from real diagnostics and runs them with the developer's privileges. No phishing, malware, or server breach is needed, and it bypasses traditional controls because every step is technically authorized. Tenet found 2,388 exposed organizations.
TrustStrike Labs