38,591 results · Feed
-
High 7.5
Panic and unbounded allocation (OOM) in the CFB/OLE2 parser on malformed input
The Compound File Binary (OLE2) parser (`src/cfb.rs`), used when reading `.xls` and `.xlsb` workbooks and when extracting VBA projects from `.xlsx`, can panic or allocate without bound on crafted input. Both are reachable through the public reader API (`open_…
-
High 7.5
Quadratic parsing of nested image wikilinks with `ENABLE_WIKILINKS`
With `Options::ENABLE_WIKILINKS`, handle_wikilink builds a wikilink's target from the raw text of its body: scan_wikilink_pipe walks it looking for |, then contains('\0') and from_replace_nuls walk it again. Once a wikilink is created, disable_all_links only …
-
Unscored —
libz-sys build script leaks environment variables in the build directory
The build script of `libz-sys` version <= 1.1.30 prints the full environment of the build, including any secrets it contains, to its stdout when compiled with `cc@1.6.0` or later on a Unix host. This affects any build that depends on `libz-sys`. Cargo saves b…
-
High 7.5
Excessive CPU and memory use on crafted GraphQL documents
Several parts of `juniper` did work that grew faster than the size of the GraphQL document. A small document could thus use seconds of CPU time or gigabytes of memory. The requests need no authentication. - `OverlappingFieldsCanBeMerged` compared all pairs of…
-
High 7.5
Stack overflow on deeply nested GraphQL documents
The parser of `juniper` called itself once for each level of nesting in a GraphQL document. The parser had no limit on the nesting depth. Deeply nested selection sets, input values, and list types thus overflowed the stack of the thread that parsed the docume…
-
High 7.5
Unbounded memory use on large request bodies
The `graphql()` and `graphql_sync()` functions of `juniper_hyper` read the whole body of a POST request into memory before they parsed the body. The functions had no limit on the size of the body. A client could thus send a large body and exhaust the memory o…
-
High 7.5CVE-2026-116690 Scanner template
MLflow: Arbitrary file read on unauthenticated MLflow servers via prompt model version source
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 …
-
Medium 6.5
MLflow CrewAI autologging records LLM provider API keys in logged trace spans
MLflow is an open source AI engineering platform for agents, large language models, and machine learning models. From 2.19.0 until 3.17.0, mlflow.crewai.autolog() records CrewAI Agent objects without consistently sanitizing nested credential-bearing values. m…
-
Critical 9.8CVE-2026-116692 Public exploit
MLflow: Remote code execution on unauthenticated MLflow servers via third-party scorer deserialization importing caller-placed artifact modules
MLflow is an open source AI engineering platform for agents, large language models, and machine learning models. From 3.12.0rc0 until 3.17.0, third-party scorer deserialization in mlflow/genai/scorers/base.py accepts request-selected descendants of THIRD_PART…
-
Medium 6.1
Check all ancestor namespace transitions when rendering raw text in html in golang.org/x/net
In childTextNodesAreLiteral, walking up the tree previously stopped at the nearest foreign ancestor once an HTML integration point was found. However, that integration point may itself be a descendant of an element that was foster-parented into a foreign name…
-
High 7.5
Avoid quadratic complexity when coalescing text nodes in html in golang.org/x/net
When coalescing adjacent text nodes in addText and fosterParent, appending to an existing TextNode's Data via string concatenation (n.Data += text) reallocated and copied the accumulated string on every token. Inputs containing many short text runs separated …
-
High 8.7
dd-trace: ReDOS in URL query string obfuscation
dd-trace is the Datadog APM client for Node.js. From 2.0.0 until 5.131.1 and 6.20.1, the default DD_TRACE_OBFUSCATION_QUERY_STRING_REGEXP can perform excessive backtracking on attacker-controlled URL query strings. The expression is applied synchronously whil…
-
High 8.5
TinyMCE Cross-Site Scripting (XSS) vulnerability using media plugin `data-mce-object` serialization
TinyMCE is an open source rich text editor. From 7.9.3 until 7.9.4 and from 8.5.1 until 8.9.3, the media plugin creates elements crafted through data-mce-object in the active document before sanitization. Those elements can execute malicious scripts during pr…
-
Medium 6.8
Model Context Protocol Python SDK: Server bearer-token authentication accepts tokens issued for a different resource server
The Model Context Protocol Python SDK is the official Python SDK for Model Context Protocol servers and clients. From 1.10.0 until 1.30.0 and 2.2.0, server-side bearer authentication configured through AuthSettings with token_verifier accepts a valid token is…
-
Medium 6.3
OpenTelemetry Python: Unbounded memory allocation in Jaeger baggage propagation
OpenTelemetry Python provides the Python API and SDK for OpenTelemetry. Prior to 1.45.0, JaegerPropagator.extract() in opentelemetry-propagator-jaeger processes inbound uberctx-* baggage values without entry, per-entry byte, or total-byte limits. Services tha…
-
Medium 5.9
Model Context Protocol Python SDK: HTTP client transports followed cross-origin redirects carrying custom headers and bodies
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_clien…
-
Medium 6.5
Model Context Protocol Python SDK: Client fetched server-chosen $ref URLs while validating tool results against output schemas
The Model Context Protocol Python SDK is the official Python SDK for Model Context Protocol servers and clients. From 1.10.0 until 1.30.0 and 2.2.0, ClientSession, ClientSessionGroup, and Client tool calls validate server-declared outputSchema documents with …
-
Low 3.1
CKEditor: Cross-site scripting (XSS) in the engine package
CKEditor 5 is a modern JavaScript rich-text editor with an MVC architecture. From 29.0.0 until 47.7.4 and 48.5.1, specially crafted markup can trigger cross-site scripting when General HTML Support is enabled and specifically configured to allow object conten…
-
High 8.1
CKEditor: Cross-site scripting (XSS) in the engine package
CKEditor 5 is a modern JavaScript rich-text editor with an MVC architecture. From 45.0.0 until 47.7.4 and 48.5.1, the engine uses a vulnerable es-toolkit version whose prototype-pollution behavior can be reached while CKEditor processes incoming style attribu…
-
Medium 5.4
Neko: Authorization bypass in legacy WebRTC input handling
Neko is a a self-hosted virtual browser that runs in Docker and uses WebRTC. From version 3.0.0 until 3.1.5, Neko deployments with legacy mode and session.implicit_hosting enabled incorrectly authorize legacy WebRTC data-channel input because LegacyIsHost() e…
-
High 8.8
Atlantis: OS command injection through PR-controlled Terraform workspace enables RCE during plan
Atlantis is a self-hosted golang application that listens for Terraform pull request events via webhooks. From version 0.4.0 until 0.47.1, Atlantis passes the pull-request-controlled workspace value from atlantis.yaml into Terraform command arguments that are…
-
High 8.6
FileBrowser Quantum: Symlink following escapes user scope allowing arbitrary file read outside the served directory
FileBrowser Quantum is a free, self-hosted, web-based file manager. Prior to 1.5.2-stable and 2.0.1-beta, file read handling in backend/pkg/indexing/indexingFiles.go resolves symbolic links through GetRealPath without confirming that the resolved target remai…
-
High 7.4
undici vulnerable to cache poisoning and cross-user disclosure via cross-origin redirects
undici is an HTTP client library for Node.js. In versions 7.0.0 through 7.30.0 and 8.0.0 through 8.12.1, when an originless Agent composes the redirect and cache interceptors in that order with a shared cache store, the cache builds its key from the original …
-
High 7.4
undici vulnerable to TLS certificate validation bypass via dropped tls options in Agent
undici is an HTTP client library for Node.js. In versions 4.0.0 through 6.29.0, 7.0.0 through 7.30.0, and 8.0.0 through 8.12.1, its Agent dispatcher passes its constructor options through a JSON-based deep clone before forwarding them to each per-origin pool,…
-
High 7.5
undici vulnerable to Denial of Service via TLS 1.2 certificate-changing renegotiation
undici is an HTTP client library for Node.js. In versions 4.2.1 through 6.29.0, 7.0.0 through 7.30.0, and 8.0.0 through 8.12.1, the client process crashes when a TLS 1.2 server renegotiates after the connection is established and presents a different, CA-trus…
-
Medium 5.9
undici WebSocket client vulnerable to Denial of Service via unbounded automatic PONG buffering
undici is an HTTP client library for Node.js. In versions 5.15.0 through 6.29.0, 7.0.0 through 7.30.0, and 8.0.0 through 8.12.1, its WebSocket client, including the bundled WebSocket exposed by Node.js, can be driven to exhaust process memory by a malicious o…
-
Medium 5.9
undici vulnerable to Denial of Service via unhandled exception in the decompress interceptor
undici is an HTTP client library for Node.js. In versions 7.15.0 through 7.30.0 and 8.0.0 through 8.12.1, its decompress interceptor forwards decompressed response data to the downstream handler from zlib stream event callbacks without catching exceptions tho…
-
High 7.5
MCP TypeScript SDK: HTTP server transports and handlers did not limit request body size or JSON-RPC batch length
### Summary In affected versions, MCP servers reachable over HTTP put no limit on the size of a POST body or on the number of messages in a JSON-RPC batch. Large requests could exhaust the server's memory, and one long batch could block it from serving anyone…
-
High 8.7
Symfony: Unbounded Recursion in the YAML Inline Lexer Lets a Single Document Crash the Process
In Symfony YAML prior to versions 5.4.54, 6.4.45, 7.4.18, and 8.1.6, Yaml\Parser::lexInlineStructure() recursed through lexInlineMapping() and lexInlineSequence() for every { and [ of an inline collection, without entering the nesting level that the rest of t…
-
High 8.7
Symfony: ReDoS in HeaderUtils::split() Reachable Through the Cache-Control Request Header
Symfony is a PHP framework for web and console applications and a set of reusable PHP components. Prior to versions 5.4.54, 6.4.46, 7.4.19, and 8.1.7, HeaderUtils::split() parsed header values with a pattern whose token alternative also matched whitespace, fo…
-
High 7.5
Go JOSE: DoS when parsing JWKs that contain RSA keys
Go JOSE provides an implementation of the Javascript Object Signing and Encryption set of standards in Go, including support for JSON Web Encryption (JWE), JSON Web Signature (JWS), and JSON Web Token (JWT) standards. Prior to 4.1.6, go-jose parses attacker-c…
-
High 7.1
CTFer.io's CTFd Terraform Provider allows a local or remote attacker to read team passwords in state as non sensitive values
CTFer.io's CTFd Terraform Provider manages CTFd resources as code. From 0.1.5 until 2.8.2, ctfd_team.password is stored as a non-sensitive value in Terraform state. A local or remote user with access to that state can read the team password without an additio…
-
High 7.8
Hydra legacy target blacklist permits unsafe PyYAML loading
## Summary Hydra's legacy target blacklist omits PyYAML's unsafe loading surfaces. If an application passes attacker-controlled targets and arguments to instantiate(), a target such as yaml.unsafe_load can execute Python code through a config-controlled YAML …
-
Medium 5.3
qs.stringify with arrayFormat comma throws TypeError on array elements that cannot be coerced to a string
### Summary `qs.stringify` with `arrayFormat: 'comma'` throws `TypeError: Cannot convert object to primitive value` when an array element cannot be coerced to a string, such as an object with an own non-callable `toString` (`{ "toString": "x" }`) or a null-pr…
-
High 7.5
undici WebSocket client vulnerable to Denial of Service via failed HTTP/2 handshake
undici is an HTTP client library for Node.js. In versions 8.11.0 and 8.11.1, its experimental WebSocket client, which is enabled by default when HTTP/2 is allowed, mishandles a failed WebSocket handshake over HTTP/2: the request is already marked completed, s…
-
High 7.4
Authlib accepts an empty HMAC key, allowing JWS/JWT signature forgery (HS256/HS384/HS512)
Authlib is a Python library which builds OAuth and OpenID Connect servers. Prior to 1.9.1, Authlib's OctKey.import_key() and HMAC signing and verification path accept zero-length key material. If an application's secret lookup can return an empty string or em…
-
Low 2.3
rustls 0.23.44 and 0.23.45 ECH Rejection Application Data Disclosure
rustls 0.23.44 and 0.23.45 contain an information disclosure vulnerability that sends application data buffered before the handshake completes to the client-facing server when Encrypted Client Hello is rejected. Because start_traffic() queues plaintext from C…
-
Low 2.3
rustls: ECH rejection can expose buffered application data to authenticated outer server before `ech_required` alert
In rustls 0.23.44 and 0.23.45, application data that a client buffers before the handshake completes is sent to the server even when the server rejects Encrypted Client Hello (ECH). When the server rejects ECH, the client completes the handshake with the oute…
-
Low 3.7
rustls before 0.23.46 Key Exchange Downgrade via Unauthenticated ServerHello
rustls before 0.23.46 contains a key exchange downgrade vulnerability that allows network attackers to poison the per-server-name key-exchange group record via an unauthenticated ServerHello. Attackers can answer one connection with a ServerHello selecting X2…
-
Low 3.7
rustls: An unauthenticated ServerHello poisons the per-name key-exchange record, so later ClientHellos carry only a classical X25519 share
## Summary The client records the server-selected key-exchange group per server name when it processes the ServerHello, before the server is authenticated. A network attacker can answer one connection attempt for a target name with a ServerHello that selects …