CVE-20XX-?
agent?

CVEs of 2026.

Every CVE published in 2026 that we hold, ranked by exploit odds.

560 CVEs · updated 2026-11-21 21:33 UTC

Reset all
  1. CVE-2026-15746

    Strands Agents is an open-source Python SDK for building and running AI agents. The strands-agents-tools package provides pre-built tools for use with the SDK, including the elasticsearch_memory tool for agent memory storage. We identified CVE-2026-15746, a server-side request forgery (SSRF) issue in the elasticsearch_memory tool. The tool exposed its connection parameters (es_url, cloud_id, api_key) as fields the large language model (LLM) could control through the tool schema. When a caller omitted the api_key parameter, the tool fell back to the operator's ELASTICSEARCH_API_KEY environment variable and sent it to whichever host the LLM specified. A crafted prompt could cause the tool to connect to a threat-actor-controlled server and disclose the operator's Elasticsearch API key in the Authorization header. We recommend you upgrade to strands-agents-tools version 0.7.0 or later. As a precautionary measure, we recommend all operators rotate their ELASTICSEARCH_API_KEY, even if there is no indication the credential was exposed.

    Disclosure Library

    6.9 MEDIUM 0.4% exploit odds Fix listed
  2. CVE-2026-19304

    IBM Langflow OSS 1.0.0 through 1.11.2 could allow a remote authenticated attacker to obtain sensitive information from internal services due to a URL parser discrepancy.

    Disclosure Library Langflow

    7.7 HIGH 0.4% exploit odds Fix listed
  3. CVE-2026-40490

    The AsyncHttpClient (AHC) library allows Java applications to easily execute HTTP requests and asynchronously process HTTP responses. When redirect following is enabled (followRedirect(true)), versions of AsyncHttpClient prior to 3.0.9 and 2.14.5 forward Authorization and Proxy-Authorization headers along with Realm credentials to arbitrary redirect targets regardless of domain, scheme, or port changes. This leaks credentials on cross-domain redirects and HTTPS-to-HTTP downgrades. Additionally, even when stripAuthorizationOnRedirect is set to true, the Realm object containing plaintext credentials is still propagated to the redirect request, causing credential re-generation for Basic and Digest authentication schemes via NettyRequestFactory. An attacker who controls a redirect target (via open redirect, DNS rebinding, or MITM on HTTP) can capture Bearer tokens, Basic auth credentials, or any other Authorization header value. The fix in versions 3.0.9 and 2.14.5 automatically strips Authorization and Proxy-Authorization headers and clears Realm credentials whenever a redirect crosses origin boundaries (different scheme, host, or port) or downgrades from HTTPS to HTTP. For users unable to upgrade, set `(stripAuthorizationOnRedirect(true))` in the client config and avoid using Realm-based authentication with redirect following enabled. Note that `(stripAuthorizationOnRedirect(true))` alone is insufficient on versions prior to 3.0.9 and 2.14.5 because the Realm bypass still re-generates credentials. Alternatively, disable redirect following (`followRedirect(false)`) and handle redirects manually with origin validation.

    Disclosure Library

    6.8 MEDIUM 0.4% exploit odds Fix listed
  4. CVE-2026-48798

    SSH.NET is a Secure Shell (SSH) library for .NET. In 2025.1.0 and earlier, ScpClient.Download(string directoryName, DirectoryInfo directoryInfo) trusts file and directory names returned by a remote SCP server and combines them with the requested local directory without containment validation, allowing a malicious, compromised, or man-in-the-middle server to use ../ sequences or absolute paths to create or overwrite files anywhere writable by the client process. This issue is fixed in version 2026.0.0.

    Disclosure Library

    7.1 HIGH 0.4% exploit odds Fix listed
  5. CVE-2026-62857

    Fedify is a TypeScript library for building federated server apps powered by ActivityPub. From version 1.2.0 through the affected 1.9, 1.10, 2.0, 2.1, 2.2, and 2.3 maintenance lines, getNodeInfo() follows an attacker-controlled links[].href value from /.well-known/nodeinfo without scheme, redirect, or private-address validation, allowing requests to loopback, link-local, cloud metadata, and private-network services and returning their response bodies. This issue is fixed in versions 1.9.13, 1.10.12, 2.0.22, 2.1.18, 2.2.7, and 2.3.2.

    Disclosure Library

    8.8 HIGH 0.4% exploit odds Fix listed
  6. CVE-2026-1801

    A flaw was found in libsoup, an HTTP client/server library. This HTTP Request Smuggling vulnerability arises from non-RFC-compliant parsing in the soup_filter_input_stream_read_line() logic, where libsoup accepts malformed chunk headers, such as lone line feed (LF) characters instead of the required carriage return and line feed (CRLF). A remote attacker can exploit this without authentication or user interaction by sending specially crafted chunked requests. This allows libsoup to parse and process multiple HTTP requests from a single network message, potentially leading to information disclosure.

    Disclosure Library Gnome Libsoup

    5.3 MEDIUM 0.4% exploit odds Fix listed
  7. CVE-2026-44243

    GitPython is a python library used to interact with Git repositories. Prior to version 3.1.48, a vulnerability in GitPython allows attackers who can supply a crafted reference path to an application using GitPython to write, overwrite, move, or delete files outside the repository’s .git directory via insufficient validation of reference paths in reference creation, rename, and delete operations. This issue has been patched in version 3.1.48.

    Disclosure Library Gitpython Project Gitpython

    7.8 HIGH 0.4% exploit odds Fix listed
  8. CVE-2026-91353

    Vulnerability in the Oracle Outside In Technology product of Oracle Fusion Middleware (component: Outside In Clean Content SDK). The supported version that is affected is 8.5.8. Easily exploitable vulnerability allows unauthenticated attacker with network access via HTTP to compromise Oracle Outside In Technology. While the vulnerability is in Oracle Outside In Technology, attacks may significantly impact additional products (scope change). Successful attacks of this vulnerability can result in unauthorized access to critical data or complete access to all Oracle Outside In Technology accessible data. CVSS 3.1 Base Score 8.6 (Confidentiality impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:N/A:N).

    Disclosure Library Oracle Outside In Technology

    8.6 HIGH 0.4% exploit odds Fix listed
  9. CVE-2026-54150

    next-video is a library for adding video to Next.js applications. Prior to 2.8.1, the GET endpoint exported by next-video/request-handler and commonly mounted at /api/video accepts an unauthenticated url query parameter, while src/utils/utils.ts isRemote() treats any value without an HTTP or HTTPS prefix as a local path. src/request-handler.ts passes that value through src/assets.ts getAssetPath() to src/config.ts loadAsset(), which appends a JSON suffix and uses fs.readFile without canonicalizing the path or verifying that it remains inside the configured video folder. A remote attacker can therefore escape the intended asset directory and read JSON files accessible to the application process, including Next.js server-action encryption material, preview-mode keys, build manifests, route metadata, absolute paths, and application video asset identifiers. Applications that do not expose the runtime request handler are not affected through this route. This issue is fixed in version 2.8.1.

    Disclosure Library

    6.9 MEDIUM 0.4% exploit odds Fix listed
  10. CVE-2026-55700

    pnpm is a package manager. From 11.3.0 until 11.5.3, `pnpm stage download` derived a local filename from registry-controlled package name and version fields. A crafted manifest could escape the selected download directory and overwrite another reachable file. The merged fix validates both fields, derives one safe filename, and verifies the final destination before writing. This vulnerability is fixed in 11.5.3.

    Disclosure Library Pnpm

    7.1 HIGH 0.4% exploit odds Fix listed
  11. CVE-2026-83167

    Vulnerability in the Oracle Application Object Library product of Oracle E-Business Suite (component: Core). Supported versions that are affected are 12.2.3-12.2.15. Easily exploitable vulnerability allows unauthenticated attacker with network access via HTTP to compromise Oracle Application Object Library. Successful attacks of this vulnerability can result in unauthorized access to critical data or complete access to all Oracle Application Object Library accessible data. CVSS 3.1 Base Score 7.5 (Confidentiality impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N).

    Disclosure Library

    7.5 HIGH 0.4% exploit odds No fix listed
  12. CVE-2026-54732

    libreoffice-convert is a Node.js module for converting office documents to different formats. Prior to 1.8.2, index.js uses the caller-controlled options.fileName value in path.join(tempDir.name, fileName) without reducing it to a base name. A filename containing ../ can escape the temporary directory because path.basename() normalization is missing and write the supplied document buffer to an arbitrary path writable by the process, including an SSH authorized_keys file, a cron configuration, or a web root. This issue is fixed in version 1.8.2.

    Disclosure Library

    6.5 MEDIUM 0.4% exploit odds Fix listed
  13. CVE-2026-53554

    SQLBot is an intelligent Text-to-SQL system based on large language models and RAG. Prior to 1.9.0, the POST /api/v1/datasource/parseExcel endpoint in backend/apps/datasource/api/datasource.py uses attacker-controlled multipart filename data when selecting where uploaded content is stored, writes the content before spreadsheet parsing and validation finish, and can transform a double-extension filename into a Python source file. An attacker able to submit a crafted multipart upload can use these behaviors to place attacker-controlled content in /opt/sqlbot/app/alembic/versions/ even when a spreadsheet parsing failure after the file write causes the endpoint to return an error. The planted file remains on disk, and subsequent SQLBot startup or migration processing causes Alembic to import the module and execute its module-level statements in the SQLBot application runtime. This issue is fixed in version 1.9.0.

    Disclosure Library

    7.3 HIGH 0.4% exploit odds Fix listed
  14. CVE-2026-120233

    A security flaw has been discovered in ag-ui-protocol ag-ui up to 0.0.58. This affects an unknown function of the file sdks/typescript/packages/client/src/agent/http.ts of the component HttpAgent. Performing a manipulation results in information disclosure. The attack can be initiated remotely. The exploit has been released to the public and may be used for attacks. The pull request to fix this issue awaits acceptance.

    Disclosure Library

    5.5 MEDIUM 0.4% exploit odds Fix listed
  15. CVE-2026-111036

    Server-side request forgery (SSRF) vulnerability in the Coupa Supplier Portal SaaS platform, specifically in the document conversion functionality available in the ‘Contracts > Templates’ section. The vulnerability is due to insufficient validation of references to external resources processed by the Aspose.Words for Java library during DOCX-to-PDF conversion. An authenticated attacker with permissions to modify templates could upload a specially crafted DOCX document containing references to external resources, which would cause the server-side application to access URLs controlled by the attacker, resources on the internal network, or local files. A successful exploit could lead to the disclosure of sensitive information, such as configuration files, system files, credentials, environment variables and endpoints for internal services.

    Disclosure Library

    8.4 HIGH 0.4% exploit odds No fix listed
  16. CVE-2026-73555

    vLLM is an inference and serving engine for large language models. Prior to 0.26.0, the validation_exception_handler in vllm/entrypoints/openai/server_utils.py converts FastAPI RequestValidationError objects with str(exc), and sanitize_message in vllm/entrypoints/utils.py does not remove traceback-style file paths, allowing unauthenticated malformed JSON requests to /v1/chat/completions, /v1/completions, /tokenize, and /detokenize to disclose the OS username, home and virtual-environment paths, Python version, internal package structure, line numbers, and endpoint handler names. This issue is fixed in version 0.26.0.

    Disclosure Library Vllm

    5.3 MEDIUM 0.4% exploit odds Fix listed
  17. CVE-2026-71557

    go-git is an extensible git implementation library written in pure Go. Prior to 5.19.2 and 6.0.0-alpha.5, reference names are not sanitized before being used to construct on-disk paths under the reference storage directory, so a maliciously crafted reference name (for example containing directory-traversal sequences) can cause go-git to write files outside the intended reference storage directory. Versions 5.19.2 and 6.0.0-alpha.5 fix the issue.

    Disclosure Library

    6.3 MEDIUM 0.4% exploit odds Fix listed
  18. CVE-2026-25121

    apko allows users to build and publish OCI container images built from apk packages. From version 0.14.8 to before 1.1.1, a path traversal vulnerability was discovered in apko's dirFS filesystem abstraction. An attacker who can supply a malicious APK package (e.g., via a compromised or typosquatted repository) could create directories or symlinks outside the intended installation root. The MkdirAll, Mkdir, and Symlink methods in pkg/apk/fs/rwosfs.go use filepath.Join() without validating that the resulting path stays within the base directory. This issue has been patched in version 1.1.1.

    Disclosure Library Chainguard Apko

    7.5 HIGH 0.4% exploit odds Fix listed
  19. CVE-2026-105283

    Generation of Error Message Containing Sensitive Information vulnerability in ash-project ash_typescript allows an unauthenticated attacker to read internal Elixir terms from an RPC error response by provoking an error the error builder has no dedicated clause for. AshTypescript.Rpc.ErrorBuilder.build_error_response/1 in lib/ash_typescript/rpc/error_builder.ex renders a pipeline error term that matches no specific clause with inspect/1 and places the result in details.error of the client-facing error. Ash exceptions take a separate path in AshTypescript.Rpc.Errors where unhandled cases are redacted behind an errorId, but these fallbacks have no redaction, so whatever the term carries reaches the HTTP or channel response. Two request paths hand it such terms: a struct-typed action argument with an invalid nested value, where cast_map_to_struct/2 in lib/ash_typescript/rpc/input_formatter.ex throws the raw Ash.Type cast error into the builder, and field selection on a generic action that returns a primitive, where format_field_type/1 inspects the full Ash.Info.Manifest.Type struct into vars.returnType. Resource module names in vars.resource are not treated as sensitive and are out of scope. This issue affects ash_typescript: from 0.1.0 before 0.19.1.

    Disclosure Library ash-project Ash Typescript

    6.3 MEDIUM 0.4% exploit odds Fix listed
  20. CVE-2026-63184

    Refitter is a tool for generating a C# REST API client using the Refit library. Prior to 2.1.1, the `refitter` dotnet tool invokes `NJsonSchema` with unsafe defaults for `$ref` resolution when generating Refit interfaces and contracts from attacker-controlled or attacker-influenced OpenAPI descriptions. Remote `http(s)` references cause requests from the developer or CI host and inline untrusted remote schema content into generated output. Absolute local file paths and relative `../` traversal outside the input directory also resolve and inline local content, permitting arbitrary local-file disclosure. The demonstrated chain stops at server-side request forgery, remote file inclusion, and local file inclusion, not code execution: Refit sets the base URL at runtime, enum values use verbatim-string doubling, and string defaults are escaped. This issue is fixed in version 2.1.1.

    Disclosure Library

    7.1 HIGH 0.4% exploit odds Fix listed
  21. CVE-2026-67529

    OpenProject is open-source, web-based project management software. Prior to 17.6.0, GET /api/v3/time_entries and GET /api/v3/cost_entries rendered _links.workPackage.title and _links.workPackage.href through associated_resource in modules/costs/lib/api/v3/time_entries/time_entry_representer.rb and modules/costs/lib/api/v3/cost_entries/cost_entry_representer.rb without checking WorkPackage.visible or view_work_packages, allowing users with view_time_entries or view_cost_entries to read private work package subjects and ids. This issue is fixed in 17.6.0.

    Disclosure Library

    4.3 MEDIUM 0.4% exploit odds Fix listed
  22. CVE-2026-79705

    A flaw was found in the buildah/copier Go package. When used outside of Buildah by a non-root caller, a crafted tar archive containing malicious symlinks can escape the target extraction directory and create files outside the intended destination. Buildah itself uses chroot hardening and is not affected.

    Disclosure Library

    4.5 MEDIUM 0.4% exploit odds No fix listed
  23. CVE-2026-41193

    FreeScout is a free self-hosted help desk and shared mailbox. Prior to version 1.8.215, FreeScout's module installation feature extracts ZIP archives without validating file paths, allowing an authenticated admin to write files arbitrarily on the server filesystem via a specially crafted ZIP. Version 1.8.215 fixes the vulnerability.

    Disclosure Library

    9.1 CRITICAL 0.4% exploit odds Fix listed
  24. CVE-2026-23942

    Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal') vulnerability in Erlang OTP (ssh_sftpd module) allows Path Traversal. This vulnerability is associated with program files lib/ssh/src/ssh_sftpd.erl and program routines ssh_sftpd:is_within_root/2. The SFTP server uses string prefix matching via lists:prefix/2 rather than proper path component validation when checking if a path is within the configured root directory. This allows authenticated users to access sibling directories that share a common name prefix with the configured root directory. For example, if root is set to /home/user1, paths like /home/user10 or /home/user1_backup would incorrectly be considered within the root. This issue affects OTP from OTP 17.0 before OTP 26.2.5.18, OTP 27.3.4.9, and OTP 28.4.1, corresponding to ssh from 3.0.1 before 5.1.4.14, 5.2.11.6, and 5.5.1. Whether OTP before OTP 17.0, corresponding to ssh before 3.0.1, is affected is unknown.

    Disclosure Library Erlang erlang\/otp

    5.3 MEDIUM 0.4% exploit odds Fix listed
  25. CVE-2026-27704

    The Dart and Flutter SDKs provide software development kits for the Dart programming language. In versions of the Dart SDK prior to 3.11.0 and the Flutter SDK prior to version 3.41.0, when the pub client (`dart pub` and `flutter pub`) extracts a package in the pub cache, a malicious package archive can have files extracted outside the destination directory in the `PUB_CACHE`. A fix has been landed in commit 26c6985c742593d081f8b58450f463a584a4203a. By normalizing the file path before writing file, the attacker can no longer traverse up via a symlink. This patch is released in Dart 3.11.0 and Flutter 3.41.0.vAll packages on pub.dev have been vetted for this vulnerability. New packages are no longer allowed to contain symlinks. The pub client itself doesn't upload symlinks, but duplicates the linked entry, and has been doing this for years. Those whose dependencies are all from pub.dev, third-party repositories trusted to not contain malicious code, or git dependencies are not affected by this vulnerability.

    Disclosure Library Dart Software Development Kit

    6.6 MEDIUM 0.4% exploit odds Fix listed
  26. CVE-2026-34726

    Copier is a library and CLI app for rendering project templates. Prior to version 9.14.1, Copier's _subdirectory setting is documented as the subdirectory to use as the template root. However, the current implementation accepts parent-directory traversal such as .. and uses it directly when selecting the template root. As a result, a template can escape its own directory and make Copier render files from the parent directory without --UNSAFE. This issue has been patched in version 9.14.1.

    Disclosure Library copier-org Copier

    4.4 MEDIUM 0.4% exploit odds Fix listed
  27. CVE-2026-111769

    The Model Context Protocol TypeScript SDK provides TypeScript client and server implementations of the Model Context Protocol. Prior to 1.32.0 in @modelcontextprotocol/sdk and from 2.0.0 until 2.3.0 in @modelcontextprotocol/client, the HTTP client transports and OAuth client requests followed redirects to a different origin while retaining custom request headers and, for HTTP 307 or 308 responses, request bodies. The affected surfaces are StreamableHTTPClientTransport, SSEClientTransport, and OAuth client functions such as auth(). The trigger is an HTTP redirect to another origin. Custom headers and mcp-session-id can be re-sent to the redirected origin. HTTP 307 and 308 redirects can also re-send the request body. OAuth request bodies can contain refresh_token and client_secret values. Current runtimes already strip the Authorization header across origins. stdio clients do not use the affected HTTP path. A malicious or compromised MCP or authorization server could therefore receive API keys, session identifiers, OAuth refresh tokens, or OAuth client secrets. Restoring unrestricted redirect following after upgrading restores the exposure. This issue is fixed in @modelcontextprotocol/sdk 1.32.0 and @modelcontextprotocol/client 2.3.0.

    Disclosure Library

    6.5 MEDIUM 0.4% exploit odds Fix listed
  28. CVE-2026-59974

    Stanza is a Stanford NLP Python library for tokenization, sentence segmentation, NER, and parsing of many human languages. Prior to 1.14.0, stanza.resources.common.unzip in stanza/resources/common.py passes downloaded model and resource archives to zipfile.ZipFile.extractall without validating member paths, and the vulnerable extraction path is reachable through stanza.download and stanza.install_corenlp. A malicious archive containing parent-directory traversal entries can write outside the intended model directory, allowing files writable by the Stanza process to be overwritten and potentially enabling code execution through modified shell configuration, SSH authorization data, Python packages, or executable scripts. This issue is fixed in version 1.14.0.

    Disclosure Library

    7.8 HIGH 0.4% exploit odds Fix listed
  29. CVE-2026-62143

    A Server-Side Request Forgery (SSRF) protection bypass existed in the html_to_markdown expansion module of misp-modules. The module attempts to prevent requests to loopback, private, link-local, and other restricted IP address ranges. However, IP addresses were compared against the blocked ranges without first normalising IPv4-mapped IPv6 addresses. An authenticated attacker able to invoke the module could supply an IPv4-mapped IPv6 address, such as: http://[::ffff:127.0.0.1]/ http://[::ffff:169.254.169.254]/ Alternatively, the attacker could use a hostname that resolves to an IPv4-mapped IPv6 address. These addresses were treated as IPv6 addresses and therefore did not match the corresponding blocked IPv4 ranges. Successful exploitation could cause the misp-modules server to connect to services available through its loopback interface, internal network, or link-local network. This could expose internal web services, administrative interfaces, or cloud instance metadata, with retrieved content potentially returned to the attacker as converted Markdown. The vulnerability has been addressed by normalising IPv4-mapped IPv6 addresses to their underlying IPv4 representation before applying the blocked-range checks. URLs without a valid hostname are now also rejected.

    Disclosure Library

    8.3 HIGH 0.4% exploit odds Fix listed
  30. CVE-2026-97323

    A vulnerability was determined in YunaiV/zhijiantianya ruoyi-vue-pro up to 2026.08. This impacts the function getOriginalFilename of the file yudao-module-mp/src/main/java/cn/iocoder/yudao/module/mp/service/material/MpMaterialServiceImpl.java of the component File Upload. Executing a manipulation can lead to path traversal. It is possible to launch the attack remotely. The exploit has been publicly disclosed and may be utilized. The vendor was contacted early about this disclosure but did not respond in any way.

    Disclosure Library

    2.1 LOW 0.4% exploit odds No fix listed
  31. CVE-2026-33872

    elixir-nodejs provides an Elixir API for calling Node.js functions. A vulnerability in versions prior to 3.1.4 results in Cross-User Data Leakage or Information Disclosure due to a race condition in the worker protocol. The lack of request-response correlation creates a "stale response" vulnerability. Because the worker does not verify which request a response belongs to, it may return the next available data in the buffer to an unrelated caller. In high-throughput environments where the library processes sensitive user data (e.g., PII, authentication tokens, or private records), a timeout or high concurrent load can cause Data A (belonging to User A) to be returned to User B. This may lead to unauthorized information disclosure that is difficult to trace, as the application may not throw an error but instead provide "valid-looking" yet entirely incorrect and private data to the wrong session. The issue is fixed in v3.1.4.

    Disclosure Library

    7.1 HIGH 0.4% exploit odds Fix listed
  32. CVE-2026-92610

    In esmf-sdk-py-aspect-model-loader versions 2.1.2 to 2.2.1, the Aspect Model loader resolves cross-file model references by extracting the third and fourth colon-separated components of any urn:samm:… namespace declared in a parsed Turtle file to build a directory to scan. A model file containing @prefix x: <urn:samm:../../../foo:bar#> or @prefix x: <urn:samm:/abs/path:bar#> causes the loader to glob and parse *.ttl files from arbitrary directories outside the intended models root, merging their contents into the returned graph.

    Disclosure Library

    9.8 CRITICAL 0.4% exploit odds No fix listed
  33. CVE-2026-62680

    Orval generates type-safe JavaScript clients in TypeScript from OpenAPI v3 and Swagger v2 specifications. Prior to 8.22.0, Orval resolves remote and local external $ref values without an allowlist or confinement to the input directory. Processing an attacker-controlled OpenAPI description can cause requests from the developer or CI host to attacker-selected or internal HTTP services, read absolute or out-of-tree local files, and inline untrusted remote schemas into generated clients. The affected code is packages/orval/src/import-specs.ts external reference loading. This issue is fixed in version 8.22.0.

    Disclosure Library

    7.1 HIGH 0.4% exploit odds Fix listed
  34. CVE-2026-50163

    oras-go is a Go library for managing OCI artifacts. Prior to 2.6.2, ensureLinkPath in content/file/utils.go:262-275 validates a hardlink target relative to the extract base but returns the unresolved target, causing os.Link("victim.secret", "<extract_base>/payload.tar.gz/evil_cwd_link") to resolve header.Linkname against the process current working directory for a Typeflag=TypeLink entry such as Name=payload.tar.gz/evil_cwd_link and Linkname="victim.secret" with io.deis.oras.content.unpack: "true", which can expose or tamper with files such as .env, .git/config, .aws/credentials, and ~/.ssh/config. This issue is fixed in version 2.6.2.

    Disclosure Library

    7.1 HIGH 0.4% exploit odds Fix listed
  35. CVE-2026-57079

    Net::BitTorrent versions before 2.1.0 for Perl write files outside the download directory via path traversal in peer-supplied metadata. Net::BitTorrent validates file path components only on the .torrent-file ingest path. The peer and magnet metadata path (_on_metadata_received, reached from the BEP09 ut_metadata extension) passes attacker-supplied file names straight to Storage::add_file and Storage::_parse_file_tree, where Path::Tiny's child() does not collapse "..". A v2 file tree key, a v1 files[].path element, or a single-file name containing ".." segments therefore resolves outside the download directory. Because the peer also controls the piece hashes and the served bytes, content verification passes, so a malicious magnet or peer writes attacker-chosen content to an attacker-chosen path on the downloading host.

    Disclosure Library

    5.3 MEDIUM 0.4% exploit odds Fix listed
  36. CVE-2026-39885

    FrontMCP is a TypeScript-first framework for the Model Context Protocol (MCP). Prior to 2.3.0, the mcp-from-openapi library uses @apidevtools/json-schema-ref-parser to dereference $ref pointers in OpenAPI specifications without configuring any URL restrictions or custom resolvers. A malicious OpenAPI specification containing $ref values pointing to internal network addresses, cloud metadata endpoints, or local files will cause the library to fetch those resources during the initialize() call. This enables Server-Side Request Forgery (SSRF) and local file read attacks when processing untrusted OpenAPI specifications. This vulnerability is fixed in 2.3.0.

    Disclosure Library Agentfront \@frontmcp\/adapters

    7.5 HIGH 0.4% exploit odds Fix listed
  37. CVE-2026-85957

    Affected Progress MarkLogic Server 11.x releases before 11.3.7 use a SAML parser configuration that can process external DTDs or entities, which could result in unintended network access or information disclosure.

    Disclosure Library

    7.5 HIGH 0.4% exploit odds No fix listed
  38. CVE-2026-3750

    A security vulnerability has been detected in ContiNew Admin up to 4.2.0. This issue affects the function URI.create of the file continew-system/src/main/java/top/continew/admin/system/factory/S3ClientFactory.java of the component Storage Management Module. The manipulation leads to server-side request forgery. The attack is possible to be carried out remotely. The exploit has been disclosed publicly and may be used. The vendor was contacted early about this disclosure but did not respond in any way.

    Disclosure Library Continew Admin

    2.0 LOW 0.4% exploit odds No fix listed
  39. CVE-2025-62616

    AutoGPT is a platform that allows users to create, deploy, and manage continuous artificial intelligence agents that automate complex workflows. Prior to autogpt-platform-beta-v0.6.34, in SendDiscordFileBlock, the third-party library aiohttp.ClientSession().get is used directly to access the URL, but the input URL is not filtered, which will cause SSRF vulnerability. This issue has been patched in autogpt-platform-beta-v0.6.34.

    Disclosure Library Agpt Autogpt Platform

    9.3 CRITICAL 0.4% exploit odds Fix listed
  40. CVE-2026-59194

    pnpm is a package manager. Prior to 10.34.4 and 11.7.0, a crafted patch entry could resolve outside the configured patches directory and cause pnpm patch-remove to delete an arbitrary reachable file. This vulnerability is fixed in 10.34.4 and 11.7.0.

    Disclosure Library Pnpm

    7.1 HIGH 0.4% exploit odds Fix listed
  41. CVE-2026-59195

    pnpm is a package manager. Prior to 10.34.4 and 11.8.0, pnpm accepts package names from the env lockfile configDependencies section and uses those names directly when creating config dependency symlinks under node_modules/.pnpm-config. A malicious repository can commit a crafted pnpm-lock.yaml whose env-lockfile document contains a traversal-shaped config dependency name. During pnpm install, pnpm installs the config dependency and creates a symlink at a path derived from that name. This vulnerability is fixed in 10.34.4 and 11.8.0.

    Disclosure Library Pnpm

    8.2 HIGH 0.4% exploit odds Fix listed
  42. CVE-2026-59196

    pnpm is a package manager. Prior to 10.34.4 and 11.7.0, a crafted lockfile alias could be joined directly under a hoisted node_modules directory. Traversal aliases could escape that directory, while reserved aliases such as .bin or .pnpm could overwrite pnpm-owned layout. This vulnerability is fixed in 10.34.4 and 11.7.0.

    Disclosure Library Pnpm

    7.1 HIGH 0.4% exploit odds Fix listed
  43. CVE-2026-116685

    The Model Context Protocol Python SDK is the official Python SDK for Model Context Protocol servers and clients. From 1.8.0 until 1.30.0 and 2.2.0, streamable_http_client, streamablehttp_client, sse_client, ClientSessionGroup, OAuth, and create_mcp_http_client client paths enable follow_redirects and follow redirects to a different origin, forwarding caller-supplied custom credential headers and, for 307 or 308 responses, the request body. An attacker who can influence a trusted endpoint's redirect destination can receive that material. This issue is fixed in versions 1.30.0 and 2.2.0.

    Disclosure Library

    5.9 MEDIUM 0.4% exploit odds Fix listed
  44. CVE-2026-45300

    The AsyncHttpClient (AHC) library allows Java applications to easily execute HTTP requests and asynchronously process HTTP responses. Versions on the 2.x branch prior to 2.15.0 and the 3.x branch prior to 3.0.10 leak `Cookie` headers to cross-origin redirect targets. When following a redirect to a different origin, the `propagatedHeaders()` method in `Redirect30xInterceptor.java` strips `Authorization` and `Proxy-Authorization` headers but does not strip the `Cookie` header, causing session cookies and other sensitive cookie values to be sent to attacker-controlled servers. Versions 2.15.0 and 3.0.10 patch the issue.

    Disclosure Library Asynchttpclient Project async-http-client

    7.4 HIGH 0.4% exploit odds Fix listed
  45. CVE-2026-103061

    Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal') vulnerability in gleam-lang gleam allows a malicious package publisher to delete a directory outside the project's build directory when a developer builds a project. A Hex package release's metadata carries an app field with the release's OTP application name. Gleam reads this field from the hex.pm release API when it downloads dependencies and uses it as the package's build directory name without validating it (hex_manifest_package in compiler-cli/src/dependencies.rs, ProjectPaths::build_directory_for_package). When a dependency fails to compile, Gleam deletes that directory with remove_dir_all. hex.pm didn't restrict the field to a directory name before 2026-10-12, so a publisher could set it to a value such as ../../../../target or an absolute path. When a project depends on such a release directly or transitively and the release fails to compile, gleam build, gleam check, gleam run, gleam test or gleam export deletes the directory the name resolves to. For a pure Gleam package no publisher code runs. This issue affects gleam: from 0.32.4 before 1.19.2.

    Disclosure Library Gleam

    5.9 MEDIUM 0.4% exploit odds Fix listed
  46. CVE-2026-112338

    Wazuh from 4.3.0 before 4.14.9 contains a server-side request forgery vulnerability in the GitHub module (wm_github.c) that trusts unvalidated Link rel="next" pagination URLs. Attackers controlling the next-page URL can make the manager send requests with the GitHub API token to attacker hosts, exposing organization repositories and settings.

    Disclosure Library

    8.2 HIGH 0.4% exploit odds Fix listed
  47. CVE-2026-7182

    Diagram's export module is vulnerable to Path Traversal in src attribute due to lack of HTML sanitization. An unauthenticated user could craft the html payload which could include local files from the server and display them in the generated pdf. This issue was fixed in version 1.1.1.

    Disclosure Library

    9.2 CRITICAL 0.4% exploit odds Fix listed
  48. CVE-2026-13343

    The UMP Stream responder library in lib/midi2/ump_stream_responder.c builds reply packets in a 16-byte struct midi_ump (uint32_t data[4]). The builders make_endpoint_info() and make_function_block_info() populate only the first two words (res.data[0] and res.data[1]) and, before this fix, declared their result as an uninitialised local (struct midi_ump res;). The remaining two words (res.data[2], res.data[3]) retain stale stack contents. Endpoint Info and Function Block Info notifications are UMP Stream messages (UMP_MT_UMP_STREAM), which are 4 words long, so the full 16-byte packet — including the two uninitialised words — is transmitted verbatim by cfg->send(). The responder is driven by attacker-supplied UMP Stream Endpoint-Discovery / Function-Block-Discovery requests via ump_stream_respond(). In the in-tree Network MIDI 2.0 server (subsys/net/lib/midi2/netmidi2.c) these requests arrive as UDP datagrams and, with the default no-authentication endpoint, a remote peer can establish a session and trigger the responses; the same library also serves USB MIDI 2.0 hosts. Each discovery request causes the device to disclose 8 bytes of its own uninitialised stack memory to the peer, and the request is freely repeatable. This is a confidentiality-only information leak (root cause is use of an uninitialised variable, CWE-457/CWE-908); the leaked words could include residual data or pointer values. There is no memory-corruption, integrity, or availability impact. The fix zero-initialises both result structs (struct midi_ump res = {0};), so the trailing words are cleared before transmission. These are the only two responder builders that left trailing words unset (send_string() already zeroes its buffer), so the leak is fully closed.

    Disclosure Library

    5.3 MEDIUM 0.4% exploit odds Fix listed
  49. CVE-2026-42315

    pyLoad is a free and open-source download manager written in Python. Prior to 0.5.0b3.dev100, when passing a folder name in the set_package_data() API function call inside the data object with key "_folder", there is no sanitization at all, allowing a user with Perms.MODIFY to specify arbitrary directories as download locations for a package. This vulnerability is fixed in 0.5.0b3.dev100.

    Disclosure Library pyload-ng Project pyload-ng

    8.1 HIGH 0.4% exploit odds Fix listed
  50. CVE-2026-118667

    LangGraph Python SDK is used to connect to running LangGraph API servers, manage assistants, threads and stream runs from Python applications. From langgraph-sdk 0.4.0 until 0.4.6, and from langgraph 1.2.3 until 1.2.14 when RemoteGraph v3 streaming uses an affected langgraph-sdk version, the synchronous and asynchronous v3 stream get_tree and _fetch_state helpers interpolate caller-controlled assistant_id and thread_id values into authenticated GET request paths without using _quote_path_param. Relative path segments can therefore be normalized into another readable path on the configured server while retaining authentication headers. This issue is fixed in langgraph-sdk 0.4.6 and langgraph 1.2.14.

    Disclosure Library

    4.3 MEDIUM 0.4% exploit odds Fix listed

CVE records published in 2026 (ids starting CVE-2026-), sorted by how likely they are to be exploited.