Total
4032 CVE
| CVE | Vendors | Products | Updated | CVSS v2 | CVSS v3 |
|---|---|---|---|---|---|
| CVE-2026-54689 | 2026-09-30 | N/A | 6.3 MEDIUM | ||
| 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. | |||||
| CVE-2026-54077 | 2026-09-30 | N/A | 7.1 HIGH | ||
| ArcadeDB is a Multi-Model DBMS. Prior to 26.6.1, the IMPORT DATABASE statement in engine/src/main/java/com/arcadedb/query/sql/parser/ImportDatabaseStatement.java did not require administrative privileges and passed its source to integration/src/main/java/com/arcadedb/integration/importer/SourceDiscovery.java without validation. An authenticated user with SQL command access through /api/v1/command or /api/v1/query can supply HTTP or HTTPS destinations to make server-side requests to internal services, or file:// paths to read files accessible to the server process and ingest the results as queryable records. The XML importer also permits DTD processing and external entities, enabling entity expansion. The root-only /api/v1/server administration endpoint is not affected. The fix requires updateSecurity permission, blocks local-network import destinations by default through arcadedb.server.security.importBlockLocalNetworks, supports the arcadedb.server.security.importAllowedLocalPaths file allow-list, and disables XML DTD processing and external entities. This issue is fixed in version 26.6.1. | |||||
| CVE-2026-55374 | 2026-09-30 | N/A | 4.8 MEDIUM | ||
| canto-saas-api is a PHP library for interacting with the Canto SaaS API. Prior to version 3.0.0, Request::buildRequestUrl() joins values returned by Request::getPathVariables() without encoding individual path segments, including the scheme and contentId values used by GetContentDetailsRequest. When a consuming application supplies an untrusted path variable value, path traversal sequences, query delimiters, or fragment delimiters can change the destination endpoint before AbstractEndpoint::sendRequest() attaches the configured authentication token. An attacker who controls that path variable value through the consuming application can cause unintended reads or writes with the configured application's privileges on the same Canto instance, but applications that pass only trusted and validated identifiers are not exploitable. This issue is fixed in version 3.0.0. | |||||
| CVE-2026-55591 | 2026-09-30 | N/A | 5.8 MEDIUM | ||
| Signal K Server is a server application that runs on a central hub in a boat. Prior to 2.28.0, makeRemoteRequest() in src/serverroutes.ts accepted attacker-controlled host, port, useTLS, and selfsignedcert parameters from the testSignalKConnection, requestAccess, and checkAccessRequest endpoints without validating the destination. When security was not configured, addAdminMiddleware() was a no-op in dummysecurity.ts, leaving all three endpoints accessible without authentication. The server could be forced to contact loopback, private, link-local, cloud metadata, or arbitrary external destinations, and selfsignedcert could disable certificate verification for outbound HTTPS requests. The checkAccessRequest endpoint also interpolated requestId into its destination path, allowing traversal to other paths on the selected host. Distinct success, connection-refused, and timeout responses enabled internal port scanning; returned response bodies enabled cloud metadata and internal-service data exfiltration; requestAccess enabled server-side POST requests with attacker-controlled JSON; and access to cluster-internal services could support lateral movement. This issue is fixed in version 2.28.0. | |||||
| CVE-2026-54637 | 2026-09-30 | N/A | N/A | ||
| Dragonfly is an open source P2P-based file distribution and image acceleration system. Prior to 2.4.4-rc.3, the scheduler's default unauthenticated v1 gRPC flow accepts attacker-controlled PeerHost.Ip and PeerHost.DownPort values through RegisterPeerTask and ReportPeerResult, storeHost copies those values into resource.Host, and handlePeerSuccess invokes Peer.DownloadTinyFile() for a TINY task. DownloadTinyFile() in scheduler/resource/standard/peer.go builds an HTTP GET request from the supplied address without destination validation, allowing a remote attacker to probe loopback, link-local, and private services and place up to TinyFileSize, 128 bytes, of a response in Task.DirectPiece for later retrieval. PeerHost.DownPort is restricted to ports 1024 through 65534, and the issue provides read SSRF rather than remote code execution. The remediation blocks loopback and link-local targets, while RFC1918 destinations remain reachable because IsGlobalUnicast accepts private ranges. This issue is fixed in 2.4.4-rc.3. | |||||
| CVE-2026-54549 | 2026-09-30 | N/A | 8.3 HIGH | ||
| Meta Ads MCP is a Model Context Protocol (MCP) server that lets AI assistants run Meta Ads. Prior to version 1.0.115, the upload_ad_image tool in meta_ads_mcp/core/ads.py passes an attacker-controlled image_url to try_multiple_download_methods() in meta_ads_mcp/core/utils.py, where httpx.AsyncClient uses follow_redirects=True and performs HTTP requests without validating the scheme, host, or resolved IP address. In a streamable-http deployment, a network caller can use any non-empty authorization value because Meta credential validation occurs after the image download, then direct the server to loopback services, private-network addresses, cloud metadata endpoints, or redirect-chained internal targets. The resulting server-side request forgery can expose internal data, invoke state-changing internal services, or disrupt reachable services. This issue is fixed in version 1.0.115. | |||||
| CVE-2026-58201 | 2026-09-30 | N/A | N/A | ||
| Lokka is a Model Context Protocol server for Microsoft 365, including Microsoft Graph and other services. Prior to 2.1.2, the Lokka-Microsoft tool in src/mcp/src/main.ts uses direct URL string concatenation to append the user-controlled path value to the management.azure.com base URL. A specially crafted path can alter URL authority parsing and cause an Azure Resource Manager bearer token to be sent to an unintended host. This issue is fixed in version 2.1.2. | |||||
| CVE-2026-73497 | 2026-09-30 | N/A | 6.5 MEDIUM | ||
| MCP Atlassian is a Model Context Protocol (MCP) server for Atlassian products (Confluence and Jira). From 0.17.0 until 0.22.0, validate_url_for_ssrf resolves the attacker-controlled X-Atlassian-Jira-Url and X-Atlassian-Confluence-Url header host once at middleware time, but the outbound request is built with the raw hostname and resolves it again at connection time with no IP pinning. An attacker-controlled DNS-rebinding name can return a public IP during validation and 169.254.169.254 or another internal IP during connection, enabling unauthenticated server-side requests to cloud metadata or internal services. The flaw spans src/mcp_atlassian/utils/urls.py, src/mcp_atlassian/servers/main.py, and src/mcp_atlassian/servers/dependencies.py; validate_url_for_ssrf returns only a verdict rather than a pinned IP, UserTokenMiddleware processes the attacker-controlled headers before fetcher creation, and the Jira and Confluence fetchers use the raw hostname. This issue is fixed in version 0.22.0. | |||||
| CVE-2026-53957 | 2026-09-30 | N/A | 7.7 HIGH | ||
| Contentful MCP Server is a Model Context Protocol server for the Contentful Management API. Prior to @contentful/mcp-server 1.7.19 and @contentful/mcp-tools 0.4.5, export_space and import_space in packages/mcp-tools/src/tools/jobs/space-to-space-migration/exportSpace.ts and packages/mcp-tools/src/tools/jobs/space-to-space-migration/importSpace.ts expose host, proxy, rawProxy, and insecure network options to LLM-controlled tool arguments and combine those options with the server's CONTENTFUL_MANAGEMENT_TOKEN. After space_to_space_migration_handler enables the migration tools, a direct MCP call or prompt injection through attacker-controlled Contentful content can redirect Contentful Management API requests and their Authorization header to an attacker-controlled host or proxy. The regular tools that use createToolClient are unaffected because those tools pin the host from server configuration. Exposure of the personal access token permits persistent out-of-band access to every Contentful space within the token's scope. This issue is fixed in @contentful/mcp-server 1.7.19 and @contentful/mcp-tools 0.4.5. | |||||
| CVE-2026-59973 | 2026-09-30 | N/A | 8.5 HIGH | ||
| FrontMCP is a TypeScript-first framework for the Model Context Protocol (MCP). From mcp-from-openapi 2.3.0 until 2.5.0 and from frontmcp and @frontmcp/adapters 1.2.1 until 1.5.0, libs/adapters/src/openapi/openapi.adapter.ts loadOpenAPISpec() forwards untrusted OpenAPI url and spec inputs and loadOptions.refResolution to OpenAPIToolGenerator.fromURL() and OpenAPIToolGenerator.fromJSON(). The external $ref guard checks parsed hostname strings without resolving addresses, pinning validated addresses, revalidating redirect targets, or normalizing IPv4-mapped IPv6. An authenticated user who can import or configure an OpenAPI specification in a hosted or multi-user deployment can use DNS-to-loopback resolution, redirect-to-loopback behavior, or IPv4-mapped IPv6 loopback forms to cause backend-origin requests to internal services. This can expose internal administrative APIs, metadata-like services, and other private network endpoints. The practical impact is lower when only a trusted local administrator can configure OpenAPI specs, and disabling external reference protocols prevents the external $ref request. This issue is fixed in mcp-from-openapi 2.5.0 and frontmcp and @frontmcp/adapters 1.5.0. | |||||
| CVE-2026-58485 | 2026-09-30 | N/A | 7.1 HIGH | ||
| 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. | |||||
| CVE-2026-63443 | 2026-09-30 | N/A | 8.3 HIGH | ||
| Coder allows organizations to provision remote development environments via Terraform. Prior to 2.29.19, 2.32.9, 2.33.10, and 2.34.4, agentConn.apiClient() follows redirects while its custom transport accepts the host from the redirected request URL when the port is the workspace agent HTTP API port 4. An authenticated user who controls a modified workspace agent and knows another online agent's UUID can derive the victim's tailnet address and redirect control-plane requests to that agent. HTTP 301, 302, and 303 redirects can redirect read requests, while HTTP 307 and 308 preserve replayable write and process-start requests. The redirected workspace agent file APIs can read or write files as the victim workspace user, and affected versions exposing the workspace agent process API can execute commands after a redirected file write, crossing workspace and tenant boundaries. This issue is fixed in versions 2.29.19, 2.32.9, 2.33.10, and 2.34.4. | |||||
| CVE-2026-59823 | 2026-09-30 | N/A | N/A | ||
| 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. | |||||
| CVE-2026-54546 | 2026-09-30 | N/A | 5.0 MEDIUM | ||
| CloudTAK is a browser-based Common Operating Picture and situational awareness tool compatible with TAK. Prior to 13.22.1, the authenticated PUT /api/basemap endpoint passes an attacker-controlled URL through importBasemapURL() in api/routes/basemap.ts to fetch(url) without resolved-address classification or redirect revalidation. BasemapProtocol.isValidURL in api/lib/interface-basemap.ts checks only the HTTP or HTTPS scheme and is not applied on the vulnerable import path. Direct internal addresses, alternate IP encodings, and redirects to internal addresses can reach cloud metadata, loopback, private, and CGNAT HTTP services. The OptionalTileJSON response reflects fields including name, attribution, and tiles[0] to the caller, making the request forgery full-read rather than blind and enabling cloud credential theft and internal service disclosure. This issue is fixed in version 13.22.1. | |||||
| CVE-2026-76820 | 2026-09-30 | N/A | 7.7 HIGH | ||
| OpenCTI is an open source platform for managing cyber threat intelligence knowledge and observables. Prior to 7.260701.0, the synchronizerFetch GraphQL query called fetchRemoteStreams after checking only that a remote stream URL used HTTP or HTTPS. The backend did not apply the ingestion deny list or reject private, loopback, and link-local destinations, allowing an authenticated account with the INGESTION capability to make OpenCTI request internal services and cloud metadata endpoints. Returned connection errors could distinguish open HTTP ports, open non-HTTP ports, and closed ports, enabling internal network scanning, while compatible endpoint responses could disclose internal data. This issue is fixed in version 7.260701.0. | |||||
| CVE-2026-54918 | 2026-09-30 | N/A | 5.3 MEDIUM | ||
| NetBox Device Type Library is a collection of community-sourced device type definitions for import into NetBox. In the affected repository revisions, NETBOX_DT_LIBRARY_URL in tests/test_configuration.py is a free-form tracked constant that an unauthenticated pull-request author can change before the validation test harness runs. During pytest collection, tests/definitions_test.py passes the value to Repo.clone_from and create_remote("upstream").fetch(), causing blind Git smart-HTTP requests to an attacker-selected host or loading attacker-controlled tests/known-*.json validation caches. The blind request cannot set arbitrary metadata-service headers or return response bodies, and this path does not execute remote Git hooks, but substituted known data can bypass slug, module, and rack uniqueness validation. This vulnerability is fixed in commit 8980c690097e92f5028c7e6df402b327d827ecd5. | |||||
| CVE-2026-102879 | 2026-09-30 | N/A | 5.0 MEDIUM | ||
| ClaraVerse through 0.3.1 contains server-side request forgery protection bypasses in the download_file and scrape_web agent tools. Authenticated users can bypass hostname validation and IPv6 transition address filtering to make the server request internal services and cloud instance metadata endpoints. | |||||
| CVE-2026-101059 | 2026-09-30 | N/A | 7.1 HIGH | ||
| utcp-http before 1.1.4 fails to validate the OAuth2 tokenUrl field from remote OpenAPI specifications, allowing attackers to redirect credential submission to arbitrary endpoints. When a victim registers an attacker-controlled OpenAPI spec and invokes a generated OAuth2-protected tool, the library POSTs the victim's client_id and client_secret to the attacker-supplied token endpoint without URL validation. | |||||
| CVE-2026-101060 | 2026-09-30 | N/A | 8.2 HIGH | ||
| python-utcp versions before 1.1.4 contain a server-side request forgery vulnerability in HttpCommunicationProtocol.call_tool that validates the initial tool URL but follows HTTP redirects without re-validating the target. Attackers controlling a tool endpoint can return a 302 redirect to internal services, allowing the UTCP client to reach cloud metadata endpoints or internal HTTP services and return their response bodies to the caller. | |||||
| CVE-2026-101061 | 2026-09-30 | N/A | 4.7 MEDIUM | ||
| utcp-gql before 1.1.1 and utcp-websocket before 1.1.1 contain server-side request forgery vulnerabilities due to incomplete application of CVE-2026-44661 fixes. The GraphQL plugin uses a vulnerable prefix check allowing bypass URLs like http://127.0.0.1.attacker.example, while the WebSocket plugin performs no URL validation despite documented security requirements. Attackers can force connections to internal services and cloud metadata endpoints by supplying malicious tool URLs in call templates, and receive configured API keys and OAuth tokens sent to attacker-controlled hosts. | |||||