| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| mcp-searxng is a Model Context Protocol server that gives AI assistants web search and URL-reading capabilities through SearXNG. Prior to 1.2.0, the web_url_read URL policy in src/url-reader.ts can be bypassed while MCP_HTTP_HARDEN is enabled and MCP_HTTP_ALLOW_PRIVATE_URLS is not enabled because redirect targets are not revalidated, 0.0.0.0 is not classified as an internal address, and IPv4-mapped IPv6 literals canonicalized to hexadecimal form are not recognized. These inputs allow an attacker-influenced tool call to make the MCP server fetch loopback or internal HTTP resources and return content from local services, private APIs, service-mesh endpoints, or cloud metadata endpoints. The separate hostname-to-private-address case addressed by the earlier partial fix is not part of these residual bypasses. This issue is fixed in version 1.2.0. |
| mcp-searxng is a Model Context Protocol server that gives AI assistants web search and URL-reading capabilities through SearXNG. Prior to 1.2.0, web_url_read passes a caller-supplied URL to the server-side fetch path while assertUrlAllowed() in src/url-reader.ts runs only when MCP_HTTP_HARDEN is enabled, even though MCP_HTTP_HARDEN is disabled by default in src/http-security.ts. In the default configuration, an attacker who influences the URL selected by a user or AI agent can make the server fetch loopback, private-network, or cloud metadata endpoint resources and return their contents into the model context. file:// URLs remain rejected, and the separate DNS-resolution and redirect-validation bypasses are outside this record. This issue is fixed in version 1.2.0. |
| mcp-searxng is a Model Context Protocol server that gives AI assistants web search and URL-reading capabilities through SearXNG. Prior to 1.7.1, web_url_read receives its caller-controlled URL through src/index.ts and validates only the literal hostname in assertUrlAllowed() within src/url-reader.ts before undiciFetch() performs operating-system DNS resolution. A public-looking attacker-controlled hostname that resolves to a private, loopback, link-local, or cloud-metadata address therefore passes the lexical check and causes the MCP server to connect to the internal destination. In the default HTTP configuration, an unauthenticated network client can use this path to read internal services, expose credentials or service tokens, and enumerate reachable internal hosts; in STDIO deployments, prompt-influenced tool selection can provide the malicious URL. Direct private IP literals are blocked, and MCP_HTTP_ALLOW_PRIVATE_URLS remains an explicit opt-out. This issue is fixed in version 1.7.1. |
| LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.83.9, an authenticated LiteLLM Proxy caller with a valid virtual key can place api_base inside the user_config request body to bypass is_request_body_safe, which blocks top-level api_base and base_url but previously did not inspect or reject user_config. Because user_config constructs the outbound router, the nested destination redirects a server-side request to an internal or external host selected by the caller and can expose endpoints the caller cannot otherwise access. This issue is fixed in version 1.83.9. |
| `@zereight/mcp-gitlab` is a Model Context Protocol server for GitLab. Starting in version 0.0.1 and prior to version 2.1.27, when the environment variable `ENABLE_DYNAMIC_API_URL=true` is set, the server reads the `X-GitLab-API-URL` HTTP request header and uses it as the base URL for all outbound GitLab API calls made within that request. The server validates that the value is a well-formed URL (`new URL(dynamicApiUrl)`) but applies no allowlist or hostname restriction. The server then attaches the victim's `Private-Token` to every outbound fetch that uses the redirected URL. Any caller who can reach the HTTP transport can set `X-GitLab-API-URL` to an attacker-controlled host. The next GitLab API call the server makes delivers the victim's token to that host. Version 2.1.27 contains a patch. |
| Nuxt OG Image generates OG Images with Vue templates in Nuxt. From 6.0.2 until 6.7.0, nuxt-og-image exposes the unauthenticated /_og/d/** route when the documented defaults security.strict = false and security.secret = "" are used, and base64url-decodes the fonts parameter through decodeOgImageParams. Attacker-controlled fonts[].path values flow through loadDefinedFonts into the font-assets/node.js binding, which performs a server-side fetch without validating the URL scheme, origin, resolved address, or redirects. This permits blind requests to loopback, private, link-local, cloud metadata, and other internal HTTP services, while differences in the outer response status and timing can reveal service reachability. Slow targets can also occupy OG image render workers for the configured fetch and render timeouts. This issue is fixed in version 6.7.0. |
| safeurl is a server-side request forgery protection library. Prior to 0.2.4, the privateNetworks list in ip.go omits the IPv6 ranges 64:ff9b:1::/48, 5f00::/16, 3fff::/20, and 100:0:0:1::/64. When an application enables IPv6 with EnableIPv6(true), an attacker-controlled destination in one of these ranges is not recognized as non-public and can pass the SSRF destination check, potentially allowing access to resources hosted within the omitted ranges. IPv6 is disabled by default, and configurations that retain EnableIPv6(false) are not exposed to this bypass. This issue is fixed in version 0.2.4. |
| gettext-converter provides gettext resource conversion utilities for JavaScript. Prior to 1.3.3, js2i18next() in lib/js2i18next.js splits nested translation keys using options.keyseparator, whose default value consists of two number signs, and uses each segment as a dynamic object key without rejecting __proto__, constructor, or prototype. When an application converts untrusted PO or i18next translation data, a __proto__ segment resolves Object.prototype as the nested write target and Object.assign writes attacker-controlled translated properties onto the process-wide prototype. The resulting prototype pollution can cause denial of service and may enable application-dependent follow-on attacks. This issue is fixed in version 1.3.3. |
| flat-to-nested converts a hierarchy from a flat representation to a nested representation. Prior to 1.1.2, FlatToNested.prototype.convert in index.js uses attacker-influenced id and parent record fields directly as keys in the plain temp and pendingChildOf objects. When parent or id is __proto__, temp[parent] can resolve to Object.prototype, and initPush() can write attacker-controlled data to the global children prototype property while existing prototype methods remain intact. Any application that passes attacker-influenced flat records to convert() can therefore expose unrelated objects to polluted inherited state, causing application-logic corruption or denial of service and potentially enabling greater impact when a downstream prototype-pollution gadget is present. The constructor and prototype strings are also unsafe inherited-key values in the same lookup design. This issue is fixed in version 1.1.2. |
| In Stapler 2107.v8dfcb_e8ed317 and earlier, except 2088.2093.vd7c3e58008a_6, included in Jenkins 2.579 and earlier, LTS 2.568.2 and earlier, Stapler does not restrict the types of objects that can be instantiated via form data binding to those compatible with the expected field type, allowing attackers with Overall/Read permission to instantiate types related to configuration for which that field type was not intended. |
| A vulnerability was determined in a2ui-project a2ui up to 0.10.6. This affects the function processMessages of the file renderers/web_core/src/v0_9/processing/message-processor.ts of the component Message Parsing. This manipulation causes dynamically-determined object attributes. The attack can be initiated remotely. The project was informed of the problem early through an issue report but has not responded yet. |
| A security flaw has been discovered in ag-ui-protocol ag-ui 0.3.0. Affected is the function urllib.request.urlopen of the file integrations/aws-strands/python/src/ag_ui_strands/utils.py of the component Multimodal Content. The manipulation of the argument Value results in server-side request forgery. The attack can be executed remotely. The patch is identified as bf0c34df34cbb4b1992bc37c9bfffe6dd54bb189. It is advisable to implement a patch to correct this issue. |
| A Server-Side Request Forgery vulnerability exists in sat_proxy.php in Zenith Satellite Tracker 1.0. The script accepts an attacker-controlled address URL parameter and passes it to curl_setopt(CURLOPT_URL) without host or scheme validation. An unauthenticated remote attacker can leverage this to make arbitrary HTTP and HTTPS requests from the server to internal networks or cloud metadata services, potentially obtaining sensitive information or pivoting to further attacks. |
| Coze Studio through 0.5.1 fails to restrict the server URL supplied when registering plugin tools, allowing authenticated users to make the backend fetch internal services. Attackers can construct plugin requests to access cloud metadata endpoints and internal services reachable only from the backend network, reading responses containing sensitive information. |
| AVideo through 29.0 contains a blind server-side request forgery vulnerability in the getHeaderContentTypeFromURL function that issues get_headers() calls guarded only by format validation. Authenticated users with canUpload permission can store attacker-chosen URLs as video links, triggering vulnerable function execution on every video watch page render to probe internal hosts using content-type oracles and timing-based detection. |
| Metabase through 0.63.18 fails to properly validate the unspecified address 0.0.0.0 in custom GeoJSON URLs, allowing unauthenticated attackers to reach loopback services. Attackers can save a malicious GeoJSON entry with 0.0.0.0 and trigger requests that return loopback service responses to unauthenticated callers. |
| HKUDS nanobot before 0.3.0 contains a server-side request forgery vulnerability in the WebFetchTool component where the _validate_url() function fails to block internal IP ranges and private addresses. Attackers can send messages instructing the bot to fetch cloud metadata endpoints, localhost services, and RFC 1918 addresses to extract IAM credentials and internal service data. |
| A vulnerability has been found in chatwoot up to 4.17.1. This impacts an unknown function of the file callbacks_controller.rb of the component Shopify OAuth. The manipulation leads to server-side request forgery. Remote exploitation of the attack is possible. The exploit has been disclosed to the public and may be used. The project was informed of the problem early through an issue report but has not responded yet. |
| An SSRF issue was discovered in OpenStack Glance before Newton. The 'copy_from' feature in the Image Service API v1 allowed an attacker to perform masked network port scans. With v1, it is possible to create images with a URL such as 'http://localhost:22'. This could then allow an attacker to enumerate internal network details while appearing masked, since the scan would appear to originate from the Glance Image service. |
| The WP Import Export Lite WordPress plugin before 3.9.33 does not properly validate URLs before requesting them during the import process, allowing users with the import capability, which administrators hold by default, to make the site issue requests to internal hosts and services and read their responses. This is an incomplete fix for CVE-2026-11397. |
