Vulnerabilities
All
My stack
Known exploited
Public exploit
Critical (9+)
Likely exploited (EPSS ≥ 50%)
Changed, 24 h
New, 24 h
Sort:
risk
modified
published
CVSS
EPSS
| CVE | CVSS | EPSS | Vendor, product | Description | Modified |
|---|---|---|---|---|---|
| CVE-2026-63801 | 8.8 high | 1.0% | linux, linux_kernel | In the Linux kernel, the following vulnerability has been resolved: tipc: fix slab-use-after-free Read in tipc_aead_decrypt_done tipc_aead_decrypt() goes straight from tipc_bearer_hold(b) to crypto_aead_decrypt(req) without taking a reference on the netns, unlike the encrypt path. When crypto_aead_decrypt() is offloaded asynchronously (e.g. the SIMD aead wrapper queuing to cryptd), the cryptd worker runs tipc_aead_decrypt_done() later. If the bearer's netns is torn down in the meantime, cleanup_net() -> tipc_exit_net() -> tipc_crypto_stop() frees the per-netns tipc_crypto, and the completion then reads it: tipc_aead_decrypt_done() dereferences aead->crypto->stats and aead->crypto->net, and tipc_crypto_rcv_complete() dereferences aead->crypto->aead[] and the node table -- reading freed memory. Decoded KASAN splat (v7.1-rc7, CONFIG_KASAN_INLINE + TIPC + TIPC_CRYPTO): BUG: KASAN: slab-use-after-free in tipc_aead_decrypt_done (net/tipc/crypto.c:999) Read of size 8 at addr ffff8881056258a8 by task kworker/u16:2/51 Workqueue: events_unbound Call Trace: tipc_aead_decrypt_done (net/tipc/crypto.c:999) process_one_work (kernel/workqueue.c:3314) worker_thread (kernel/workqueue.c:3397 kernel/workqueue.c:3478) kthread (kernel/kthread.c:436) ret_from_fork (arch/x86/kernel/process.c:158) ret_from_fork_asm (arch/x86/entry/entry_64.S:245) Allocated by task 169: __kasan_kmalloc (mm/kasan/common.c:398 mm/kasan/common.c:415) tipc_crypto_start (net/tipc/crypto.c:1502) tipc_init_net (net/tipc/core.c:72) ops_init (net/core/net_namespace.c:137) setup_net (net/core/net_namespace.c:446) copy_net_ns (net/core/net_namespace.c:579) create_new_namespaces (kernel/nsproxy.c:132) __x64_sys_unshare (kernel/fork.c:3316) do_syscall_64 (arch/x86/entry/syscall_64.c:63) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:121) Freed by task 8: kfree (mm/slub.c:6566) tipc_exit_net (net/tipc/core.c:119) cleanup_net (net/core/net_namespace.c:704) process_one_work (kernel/workqueue.c:3314) kthread (kernel/kthread.c:436) This is the same class of bug that commit e279024617134 ("net/tipc: fix slab-use-after-free Read in tipc_aead_encrypt_done") fixed for the encrypt side. The encrypt path takes maybe_get_net(aead->crypto->net) before crypto_aead_encrypt() and drops it with put_net() on the synchronous return paths and in tipc_aead_encrypt_done(); the -EINPROGRESS/-EBUSY return keeps the reference for the async callback to release. The decrypt path was left without the equivalent guard. Mirror the encrypt-side fix on the decrypt path: take a net reference before crypto_aead_decrypt() (failing with -ENODEV and the matching bearer put if it cannot be acquired), keep it across the -EINPROGRESS/-EBUSY async return, and drop it with put_net() on the synchronous success/error return and at the end of tipc_aead_decrypt_done(). Reproduced under KASAN on v7.1-rc7: a UDP bearer with a cluster key is flooded with crafted encrypted frames from an unknown peer (driving the cluster-key decrypt path) while the bearer's netns is repeatedly torn down. The completion must run asynchronously to outlive tipc_crypto_stop(); on x86 the stock aesni gcm(aes) now decrypts synchronously, so the async path was exercised via cryptd offload. The unguarded aead->crypto dereference in tipc_aead_decrypt_done() is the unpatched upstream path; tipc_aead_decrypt() still lacks maybe_get_net(aead->crypto->net), so the completion can outlive the free on any config where crypto_aead_decrypt() goes async. Found by 0sec automated security-research tooling (https://0sec.ai). | 19 Oct |
| CVE-2025-61591 | 8.8 high | 1.0% | anysphere, cursor | Cursor is a code editor built for programming with AI. In versions 1.7 and below, when MCP uses OAuth authentication with an untrusted MCP server, an attacker can impersonate a malicious MCP server and return crafted, maliciously injected commands during the interaction process, leading to command injection and potential remote code execution. If chained with an untrusted MCP service via OAuth, this command injection vulnerability could allow arbitrary code execution on the host by the agent. This can then be used to directly compromise the system by executing malicious commands with full user privileges. This issue does not currently have a fixed release version, but there is a patch, 2025.09.17-25b418f. | 19 Oct |
| CVE-2025-64109 | 8.8 high | 1.0% | no CPE yet | Cursor is a code editor built for programming with AI. In versions and below, a vulnerability in the Cursor CLI Beta allowed an attacker to achieve remote code execution through the MCP (Model Context Protocol) server mechanism by uploading a malicious MCP configuration in .cursor/mcp.json file in a GitHub repository. Once a victim clones the project and opens it using Cursor CLI, the command to run the malicious MCP server is immediately executed without any warning, leading to potential code execution as soon as the command runs. This issue is fixed in version 2025.09.17-25b418f. | 19 Oct |
| CVE-2026-62675 | 8.8 high | 1.0% | no CPE yet | Omnigent is an open-source AI agent framework and meta-harness for orchestrating coding agents. Prior to 0.3.0, multipart POST /v1/sessions accepts an authenticated user's agent bundle and omnigent/server/bundles.py validate_agent_bundle does not reject a tools..callable dotted Python path. omnigent/runner/tool_dispatch.py _resolve_spec_callable imports the specified module and _execute_spec_callable_tool invokes the resolved function, allowing a bundle to select subprocess.check_output and execute a local command with the runner process permissions. This can expose runner files, environment variables, credentials, workspace data, internal services, and availability without administrator access. This issue is fixed in version 0.3.0. | 14 Oct |
| CVE-2026-62677 | 8.8 high | 1.0% | no CPE yet | Omnigent is an open-source AI agent framework and meta-harness for orchestrating coding agents. Prior to 0.3.0, an authenticated user can upload a session-scoped agent bundle with an absolute or traversal-containing os_env.cwd value because omnigent/spec/parser.py stores the value verbatim and omnigent/spec/validator.py does not constrain it. On a runner where OMNIGENT_RUNNER_WORKSPACE is unset, omnigent/runner/resource_registry.py preserves the attacker-controlled path and omnigent/inner/os_env.py uses the resolved path as the environment root and copytree source. The _assert_within_cwd check then treats that attacker-selected root as trusted, allowing sys_os_read, write, edit, and shell tools to access runner files and environment secrets outside the intended workspace. This issue is fixed in version 0.3.0. | 14 Oct |
| CVE-2026-42856 | 8.7 high | 1.0% | no CPE yet | Network-AI is a TypeScript/Node.js multi-agent orchestrator. Prior to 5.1.3, the MCP HTTP transport accepts JSON-RPC tools/call requests with no authentication, session, origin, or token check, and dispatches them directly to the orchestrator's tool registry. The default bind address is 0.0.0.0. As a result, any party with network reachability to the service can enumerate and invoke privileged management tools. This vulnerability is fixed in 5.1.3. | Mon 26 Oct |
| CVE-2026-53977 | 8.7 high | 1.0% | no CPE yet | OpenChamber 1.11.7 contains an authentication bypass vulnerability that allows unauthenticated remote attackers to terminate the server process by sending a POST request to the /api/system/shutdown endpoint, which is registered before the authentication middleware in the Express route handler chain. Attackers can exploit the route registration order in bootstrap-runtime.js to reach the shutdown handler before auth middleware executes, causing denial of service to all active AI coding sessions and locking out legitimate remote users regardless of whether UI_PASSWORD is configured. | Sat 24 Oct |
| CVE-2026-73614 | 8.7 high | 1.0% | no CPE yet | Network-AI ClaudeHookBridge before 5.15.1 truncates the target string to 500 characters before evaluating denyPatterns, while Claude Code executes the full untruncated command. Attackers can position dangerous content past byte 500 in a Bash command field to bypass the operator's hard-deny list and execute arbitrary commands. | Sat 24 Oct |
| CVE-2026-73615 | 8.7 high | 1.0% | no CPE yet | Network-AI versions before 5.15.1 contain a security matcher bypass vulnerability where SandboxPolicy evaluates raw command strings with quotes preserved while the executor tokenizes commands by stripping quotes before execution. Attackers can craft quoted commands that evade blocklist checks and approval gates while the executor runs the identical unquoted dangerous argv. | Sat 24 Oct |
| CVE-2026-73298 | 8.7 high | 1.0% | no CPE yet | The Microsoft Container Migration Solution Accelerator is a multi-service application that provides a multi-agent, AI-driven migration solution for moving container service configurations to Azure Kubernetes Service. In version 2.1.2 and earlier, a security vulnerability was identified in the Container Migration Solution Accelerator, specifically an authenticated IDOR (Insecure Direct Object Reference) that allows users to read, write, and delete processes belonging to other authenticated users. The issue affects multiple API endpoints, where ownership checks are missing, enabling unauthorized access and modification of migration data across users within the same organization. The vulnerability is present in both process and file management APIs, and the application relies on Entra ID authentication but lacks proper authorization controls between users. Authenticated users are able to access, modify, and delete processes and files belonging to other users without proper authorization checks. | Sat 24 Oct |
| CVE-2026-75859 | 8.7 high | 1.0% | no CPE yet | CodeWhale versions before 0.8.64 fail to validate file paths in the project config instructions field, allowing attackers to read arbitrary files on the victim's system. A malicious .codewhale/config.toml file in a cloned repository can specify paths outside the workspace that are read and injected into the AI system prompt for exfiltration. | 22 Oct |
| CVE-2026-57206 | 8.6 high | 1.0% | no CPE yet | SimpleChat is a secure AI conversation application with personal and group workspaces for document-grounded interactions. Prior to 0.241.206, several plugin validation routes in application/single_app/plugin_validation_endpoint.py, including `POST /api/admin/plugins/test-instantiation`, `GET /api/admin/plugins/health-check/<plugin_name>`, `POST /api/admin/plugins/repair/<plugin_name>`, and `POST /api/plugins/validate`, relied on @swagger_route(security=get_auth_security()) documentation without enforcing @login_required, @user_required, or @admin_required at runtime, allowing unauthenticated or unauthorized clients to invoke plugin validation, health, and repair behavior. This issue is fixed in version 0.241.206. | Sat 24 Oct |
| CVE-2026-117558 | 8.5 high | 1.0% | no CPE yet | The ai-sdlc plugin MCP server through 0.31.2 contains an OS command injection vulnerability that allows attackers to execute commands by supplying shell metacharacters in the repo argument. The unvalidated repo value is interpolated into gh issue view or gh pr view commands run via execSync(), letting prompt-injected agents execute commands as the MCP server user. | Tue 27 Oct |
| CVE-2026-44688 | 8.4 high | 1.0% | eclipse, theia | In Eclipse Theia versions prior to 1.71.0, the AI chat agent processed workspace file and directory names as part of its prompt context without distinguishing them from system instructions. An attacker could craft a malicious repository with adversarial directory or file names that, when analyzed by the AI agent, would cause the agent to follow attacker-controlled instructions (indirect prompt injection). Combined with other AI chat features available in untrusted workspaces, this enabled attack chains leading to data exfiltration via Markdown image rendering or arbitrary command execution via task definitions. | Mon 26 Oct |
| CVE-2026-46580 | 8.4 high | 1.0% | eclipse, theia | In Eclipse Theia versions prior to 1.71.0, files matching the pattern .prompts/*.prompttemplate in a workspace were automatically loaded and could override or extend the AI agent's system prompts. An attacker could craft a malicious repository containing prompt template files that, when the workspace was opened in Theia, replaced the AI's system instructions with attacker-controlled content (indirect prompt injection). Combined with other AI chat features available in untrusted workspaces, this enabled attack chains leading to data exfiltration via Markdown image rendering or arbitrary command execution via task definitions. | Mon 26 Oct |
| CVE-2026-56675 | 8.3 high | 1.0% | no CPE yet | 9Router is an AI router & token saver. Prior to 0.5.2, 9router treats loopback requests as trusted and allows /v1/* access without an API key, so a same-host reverse proxy that forwards public traffic to the backend through 127.0.0.1 causes src/dashboardGuard.js to misclassify external requests as local. A remote unauthenticated attacker can access /v1 APIs such as /v1/models and may abuse configured upstream provider credentials through /v1 proxy endpoints depending on enabled providers. This issue is fixed in version 0.5.2. | Sun 25 Oct |
| CVE-2026-41432 | 8.2 high | 1.0% | newapi, new_api | New API is a large language mode (LLM) gateway and artificial intelligence (AI) asset management system. Prior to version 0.12.10, a vulnerability exists in the Stripe webhook handler that allows an unauthenticated attacker to forge webhook events and credit arbitrary quota to their account without making any payment. This issue has been patched in version 0.12.10. | Mon 26 Oct |
| CVE-2026-44016 | 8.2 high | 1.0% | docling | Docling simplifies document processing by parsing diverse formats and providing integrations with the generative AI ecosystem. FIn versions >= 2.82.0, < 2.91.0, if the HTML backend was explicitly configured for rendering (rendering option by default deactivated), then the Playwright-based rendering feature could allow JavaScript execution and unrestricted network access when processing untrusted HTML documents. An attacker could craft malicious HTML that executes arbitrary JavaScript in the rendering context or makes unauthorized network requests to internal services, potentially leading to SSRF attacks, data exfiltration, or remote code execution in the rendering environment. This vulnerability is fixed in 2.91.0. | Fri 23 Oct |
| CVE-2026-65842 | 8.2 high | 1.0% | no CPE yet | Plate is a rich-text editor with AI and shadcn/ui. Prior to 53.3.2, @platejs/docx-io fetches remote image URLs while converting attacker-controlled HTML through htmlToDocxBlob in a server-side or privileged environment. The converter can make requests to internal network resources and include the fetched image bytes in the generated DOCX, allowing server-side request forgery with response disclosure. Applications can also incur resource consumption from attacker-selected remote responses. This issue is fixed in version 53.3.2. | 18 Oct |
| CVE-2026-42075 | 8.1 high | 1.0% | no CPE yet | Evolver is a GEP-powered self-evolving engine for AI agents. Prior to version 1.69.3, a path traversal vulnerability in the skill download (fetch) command allows attackers to write files to arbitrary locations on the filesystem. The --out= flag accepts user-provided paths without validation, enabling directory traversal attacks that can overwrite critical system files or create files in sensitive location. This issue has been patched in version 1.69.3. | Mon 26 Oct |
| CVE-2026-50758 | 8.1 high | 1.0% | no CPE yet | Cross Site Scripting vulnerability in DayuanJiang next-ai-draw-io 0.4.13 allows a remote attacker to execute arbitrary code via the mcp parameter | Sat 24 Oct |
| CVE-2026-17496 | 8.1 high | 1.0% | notegen | NoteGen before 0.32.0 renders AI chat responses with markdown-it configured with html:true and injects the result into the DOM via dangerouslySetInnerHTML in chat-preview, without HTML sanitization and with CSP set to null. Attacker-controlled content that reaches the model prompt (for example a malicious skill REFERENCE.md that instructs the model to emit HTML) can cause the model response to include executable markup such as an img onerror handler. When the user views the chat response, that markup runs as JavaScript in the privileged Tauri webview, enabling arbitrary script execution in the application context (cross-site scripting). | Sat 24 Oct |
| CVE-2026-19942 | 8.1 high | 1.0% | no CPE yet | The Atarim – AI Agency for WordPress: Edit Pages, Fix Code, Update Plugins, SEO & Client Feedback plugin for WordPress is vulnerable to arbitrary file deletion due to insufficient file path validation in the AVCF_Abilities_Media::register (replace-media-file execute_callback) function in all versions up to, and including, 5.1.1. This makes it possible for authenticated attackers, with author-level access and above, to delete arbitrary files on the server, which can easily lead to remote code execution when the right file is deleted (such as wp-config.php). This is exploitable by first using the atarim/update-post-field ability to overwrite the _wp_attached_file meta of an attacker-owned attachment with a directory-traversal path, then invoking atarim/replace-media-file to cause get_attached_file() to resolve and unlink the targeted file. | 22 Oct |
| CVE-2025-65946 | 8.1 high | 1.0% | roocode, roo_code | Roo Code is an AI-powered autonomous coding agent that lives in users' editors. Prior to version 3.26.7, Due to an error in validation it was possible for Roo to automatically execute commands that did not match the allow list prefixes. This issue has been patched in version 3.26.7. | 18 Oct |
| CVE-2026-110455 | 8.1 high | 1.0% | no CPE yet | A malicious actor with access to the network and under certain conditions could exploit an Improper Input Validation vulnerability found in UniFi Protect AI Key to execute a Command Injection on the device. | 15 Oct |
| CVE-2026-17092 | 8.1 high | 1.0% | no CPE yet | The Directorist: AI-Powered Business Directory, Listings & Classified Ads plugin for WordPress is vulnerable to authorization bypass in all versions up to, and including, 8.10.2.1. This is due to the plugin not properly verifying that a user is authorized to perform an action. This makes it possible for authenticated attackers, with subscriber-level access and above, to delete arbitrary business directory listings, including those owned by other users and posts in publish, draft, or private status. The nonces guarding the endpoint do not constitute an authorization barrier because both directorist_nonce and atbdp_nonce_js are exposed to all logged-in users via wp_localize_script. | 14 Oct |
| CVE-2025-10488 | 8.1 high | 1.0% | no CPE yet | The Directorist: AI-Powered Business Directory Plugin with Classified Ads Listings plugin for WordPress is vulnerable to arbitrary file move due to insufficient file path validation in the add_listing_action AJAX action in all versions up to, and including, 8.4.8. This makes it possible for unauthenticated attackers to move arbitrary files on the server, which can easily lead to remote code execution when the right file is moved (such as wp-config.php). | 8 Oct |
| CVE-2026-15218 | 7.9 high | 1.0% | no CPE yet | A flaw was found in the maas-api and maas-controller ServiceAccounts within Red Hat OpenShift AI. These ServiceAccounts are granted cluster-wide permissions that exceed their operational requirements. An attacker who compromises the identity of these ServiceAccounts, either through a remote code execution vulnerability or by creating a malicious pod in the same namespace, could exploit these excessive permissions. This could lead to full cluster administrator privileges through the creation of new ClusterRoleBindings or the disclosure of sensitive information by accessing all secrets across the cluster. | 19 Oct |
| CVE-2026-4775 | 7.8 high | 1.0% | libtiff, debian, debian_linux, ai_inference_server | A flaw was found in the libtiff library. A remote attacker could exploit a signed integer overflow vulnerability in the putcontig8bitYCbCr44tile function by providing a specially crafted TIFF file. This flaw can lead to an out-of-bounds heap write due to incorrect memory pointer calculations, potentially causing a denial of service (application crash) or arbitrary code execution. | Mon 26 Oct |
| CVE-2026-50757 | 7.8 high | 1.0% | no CPE yet | Directory Traversal vulnerability in DayuanJiang next-ai-draw-io 0.4.13 allowsa remote attacker to execute arbitrary code via the nex-ai-draw-io/mcp-server | Sat 24 Oct |
| CVE-2026-69923 | 7.8 high | 1.0% | microsoft, windows_11_24h2, windows_11_25h2, windows_11_26h1 | Uncontrolled search path element in Windows AI allows an authorized attacker to elevate privileges locally. | 16 Oct |
| CVE-2026-108456 | 7.7 high | 1.0% | no CPE yet | Server-side request forgery (ssrf) in Azure AI Foundry allows an authorized attacker to elevate privileges over a network. | Mon 26 Oct |
| CVE-2026-116690 | 7.5 high | 1.0% | no CPE yet | MLflow is an open source AI engineering platform for agents, large language models, and machine learning models. From 3.10.0 until 3.17.0, the prompt branch in _create_model_version() accepts schemeless relative sources because path validation runs only when a source has a URL scheme, and mlflow.prompt.is_prompt is interpreted by tag presence rather than a true value. A remote caller can store a server-local path as the model version artifact location and use /api/2.0/mlflow/model-versions/get-artifact to read files accessible to the MLflow server process. Default unauthenticated servers require no privileges, while basic-auth deployments are exposed to authenticated users who can create registered models. This issue is fixed in version 3.17.0. | Wed 28 Oct |
| CVE-2026-85917 | 7.5 high | 1.0% | microsoft, azure ai foundry | Server-side request forgery (ssrf) in Azure AI Foundry allows an unauthorized attacker to elevate privileges over a network. | Mon 26 Oct |
| CVE-2026-55446 | 7.5 high | 1.0% | langflow | Langflow is a tool for building and deploying AI-powered agents and workflows. Prior to 1.0.19, an attacker can send a /api/v1/files/upload/ request without any authentication token/cookies and abuse a very long multipart form boundary to make the langflow app unusable for all users for an indefinite amount of time. This vulnerability is fixed in 1.0.19. | Sun 25 Oct |
| CVE-2026-44017 | 7.5 high | 1.0% | docling | Docling simplifies document processing by parsing diverse formats and providing integrations with the generative AI ecosystem. Prior to 2.91.0, the EasyOCR model download functionality extracted ZIP archives without validating member paths, enabling Zip Slip attacks. If an attacker could compromise the model download source (via supply chain attack, DNS spoofing, or MITM), they could write arbitrary files to any location writable by the process, potentially achieving remote code execution by overwriting Python files or system binaries, persistent backdoors by modifying startup scripts or SSH keys, and data corruption or system compromise. This vulnerability is fixed in 2.91.0. | Sun 25 Oct |
| CVE-2026-44020 | 7.5 high | 1.0% | docling | Docling simplifies document processing by parsing diverse formats and providing integrations with the generative AI ecosystem. From 2.13.0 until 2.74.0, the USPTO patent XML parser used the standard xml.sax.parseString() without protection against XML External Entity (XXE) attacks. An attacker could craft malicious USPTO patent XML files with external entity references that could read arbitrary files from the server filesystem, perform Server-Side Request Forgery (SSRF) attacks, or cause denial of service through entity expansion (Billion Laughs attack). The vulnerability affects three USPTO patent format parsers: ICE (v4.x), Grant v2.5, and Application v1.x. This vulnerability is fixed in 2.74.0. | Sun 25 Oct |
| CVE-2026-49987 | 7.5 high | 1.0% | yamadashy, repomix | Repomix is a tool that packs repositories into AI-friendly files. Prior to 1.14.1, src/core/git/gitCommand.ts execGitShallowClone passes the --remote-branch value directly to git fetch and git checkout without validation or --end-of-options, allowing --upload-pack or other Git option injection that bypasses validateGitUrl() dangerous parameter checks and can execute commands through local or SSH-style transports. This issue is fixed in version 1.14.1. | Sat 24 Oct |
| CVE-2026-15322 | 7.5 high | 1.0% | ibm, engineering_ai_hub | IBM Engineering AI Hub 1.0.0, 1.1.0, and 1.2.0 could allow a remote attacker to obtain sensitive information due to the exposure of session tokens in URLs. | Sat 24 Oct |
| CVE-2026-50756 | 7.5 high | 1.0% | no CPE yet | An issue in DayuanJiang next-ai-draw-io 0.4.13 allows a remote attacker to obtain sensitive information via the x-ai-provider component | Sat 24 Oct |
| CVE-2025-40187 | 7.5 high | 1.0% | no CPE yet | In the Linux kernel, the following vulnerability has been resolved: net/sctp: fix a null dereference in sctp_disposition sctp_sf_do_5_1D_ce() If new_asoc->peer.adaptation_ind=0 and sctp_ulpevent_make_authkey=0 and sctp_ulpevent_make_authkey() returns 0, then the variable ai_ev remains zero and the zero will be dereferenced in the sctp_ulpevent_free() function. | 19 Oct |
| CVE-2025-61590 | 7.5 high | 1.0% | anysphere, cursor | Cursor is a code editor built for programming with AI. Versions 1.6 and below are vulnerable to Remote Code Execution (RCE) attacks through Visual Studio Code Workspaces. Workspaces allow users to open more than a single folder and save specific settings (pretty similar to .vscode/settings.json) for the folders / project. An untitled workspace is automatically created by VS Code (untitled.code-workspace), which contains all the folders and workspace settings from the user's current session, opening up an entire new attack vector if the user has a .code-workspace file in path (either untitled created automatically or a saved one). If an attacker is able to hijack the chat context of the victim (such as via a compromised MCP server), they can use prompt injection to make the Cursor Agent write into this file and modify the workspace. This leads to a bypass of CVE-2025-54130 which can lead to RCE by writing to the settings section. This issue is fixed in version 1.7. | 19 Oct |
| CVE-2025-62039 | 7.5 high | 1.0% | no CPE yet | Insertion of Sensitive Information Into Sent Data vulnerability in Ays Pro AI ChatBot with ChatGPT and Content Generator by AYS ays-chatgpt-assistant allows Retrieve Embedded Sensitive Data.This issue affects AI ChatBot with ChatGPT and Content Generator by AYS: from n/a through <= 2.6.6. | 19 Oct |
| CVE-2026-59279 | 7.5 high | 1.0% | vmware, spring_ai | The MCP Streamable HTTP server transport (WebFlux and WebMvc variants) does not place any limit on the number of sessions it retains, and by default does not require clients to be authenticated. As a result, a remote attacker can cause the server to accumulate an unbounded number of sessions over time, gradually exhausting available memory and ultimately causing a Denial of Service that affects all legitimate clients. Affected versions: Spring AI: 2.0.0 | 16 Oct |
| CVE-2025-66909 | 7.5 high | 1.0% | turms-im, turms | Turms AI-Serving module v0.10.0-SNAPSHOT and earlier contains an image decompression bomb denial of service vulnerability. The ExtendedOpenCVImage class in ai/djl/opencv/ExtendedOpenCVImage.java loads images using OpenCV's imread() function without validating dimensions or pixel count before decompression. An attacker can upload a specially crafted compressed image file (e.g., PNG) that is small when compressed but expands to gigabytes of memory when loaded. This causes immediate memory exhaustion, OutOfMemoryError, and service crash. No authentication is required if the OCR service is publicly accessible. Multiple requests can completely deny service availability. | 13 Oct |
| CVE-2026-45539 | 7.4 high | 1.0% | no CPE yet | Microsoft APM is an open-source, community-driven dependency manager for AI agents. From 0.5.4 to 0.12.4, two primitive integrators in apm-cli enumerate package files with bare Path.glob() / Path.rglob() calls and read each match with Path.read_text(), transparently following symbolic links. A symlink committed inside a remote APM dependency under .apm/prompts/<x>.prompt.md or .apm/agents/<x>.agent.md is preserved verbatim into apm_modules/ on clone and then dereferenced during integration, with the resolved content written as a regular file into the project's deploy directories. The package content_hash, the pre-deploy SecurityGate scan, and apm audit do not flag this. The deploy roots are not added to the auto-generated .gitignore, so the resulting files are staged by git add by default. This vulnerability is fixed in 0.13.0. | Mon 26 Oct |
| CVE-2026-48787 | 7.4 high | 1.0% | no CPE yet | gin-vue-admin is an AI-assisted basic development platform. In version 2.9.1, an authenticated attacker with access to the code-generation feature and MCP management interface can exploit this vulnerability by injecting attacker-controlled Go source code through POST /autoCode/addFunc, and then invoking POST /autoCode/mcpStart to trigger a rebuild and restart of the standalone MCP service. This allows arbitrary operating system commands to be executed on the server with the privileges of the application process. Successful exploitation may lead to remote code execution (RCE), modification of backend source code or runtime logic, deployment of persistent backdoors, access to or manipulation of application data and configuration, and further impact on local resources running under the same service account or privilege context. The risk is highest in deployments that retain the source tree, allow writes to source files, and support local build or startup of standalone MCP components. In environments using binary-only releases, read-only filesystems, or with local build capabilities removed, the exploitability of the full attack chain is significantly reduced. However, once the online code-generation capability and MCP-hosted startup workflow are enabled, the overall security impact may reach high to critical severity. As of time of publication, it is unknown if a patched version is available. As a workaround, enforce strict allowlist validation on path- and identifier-related fields such as `humpPackageName`, `packageName`, `FuncName`, and `Router`, and only permit safe identifier formats. | Sun 25 Oct |
| CVE-2026-96893 | 7.3 high | 1.0% | gitlab | GitLab has remediated an issue in GitLab EE affecting all versions from 19.1 before 19.2.8, 19.3 before 19.3.5, and 19.4 before 19.4.2 that, under certain conditions, could have allowed an authenticated user to inject specially crafted content into AI agent instructions, potentially causing unintended command execution attributed to other users, due to improper neutralization of special elements in AI conversation context construction. | 14 Oct |
| CVE-2026-65015 | 7.2 high | 1.0% | n8n | n8n versions before 2.30.1 contain a privilege escalation vulnerability in the AI Agents feature where the node-execution tool lacks proper authorization checks. A Project Viewer user can escalate privileges by chatting with an agent that has node tools enabled, executing arbitrary nodes and accessing credential secrets without proper authorization verification. | Sat 24 Oct |
| CVE-2026-39979 | 6.9 medium | 1.0% | jqlang, redhat, jq, ai_inference_server, ansible_automation_platform | jq is a command-line JSON processor. In commits before 2f09060afab23fe9390cce7cb860b10416e1bf5f, the jv_parse_sized() API in libjq accepts a counted buffer with an explicit length parameter, but its error-handling path formats the input buffer using %s in jv_string_fmt(), which reads until a NUL terminator is found rather than respecting the caller-supplied length. This means that when malformed JSON is passed in a non-NUL-terminated buffer, the error construction logic performs an out-of-bounds read past the end of the buffer. The vulnerability is reachable by any libjq consumer calling jv_parse_sized() with untrusted input, and depending on memory layout, can result in memory disclosure or process termination. The issue has been patched in commit 2f09060afab23fe9390cce7cb860b10416e1bf5f. | Mon 26 Oct |
Download these 718 as CSV · subscribe to this list.
EPSS scores: 2026-10-29. NVD cursor: 2026-10-29 00:22. The News page shows only notable CVEs (KEV, your stack, new criticals, EPSS over 50%, or active exploitation per SSVC); this table has all of them.