ghsa

### Impact `API_URLS` is utilizing HTTP instead of HTTPS for communication that can lead to issues like Eavesdropping, Data Tampering, Unauthorized Data Access & MITM Attacks. ### References [ISSUE](https://github.com/ARPSyndicate/puncia/issues/8) [PATCH](https://github.com/ARPSyndicate/puncia/commit/033f3b68126eabbb2040ce16e2c3a2ce17437fbd#diff-3ec6c2de51e702726b23c452e3f4a899f6f4253af9fbf5be7254a5c1407ab526)

### Impact The server allow to create any user who can trigger a pipeline run malicious workflows: - Those workflows can either lead to a host takeover that runs the agent executing the workflow. - Or allow to extract the secrets who would be normally provided to the plugins who's entrypoint are overwritten. ### Patches https://github.com/woodpecker-ci/woodpecker/pull/3933 ### Workarounds _Is there a way for users to fix or remediate the vulnerability without upgrading?_ **Enable the "gated" repo feature and review each change upfront** ### References - https://github.com/woodpecker-ci/woodpecker/pull/3933 - https://github.com/woodpecker-ci/woodpecker-security/pull/11 - https://github.com/woodpecker-ci/woodpecker-security/issues/8 (info will be published later at https://github.com/woodpecker-ci/woodpecker/issues/3924) - https://github.com/woodpecker-ci/woodpecker-security/issues/9 (info will be published later at https://github.com/woodpecker-ci/woodpecker/issues/3924) - https://gi...

### Impact The server allow to create any user who can trigger a pipeline run malicious workflows: - Those workflows can either lead to a host takeover that runs the agent executing the workflow. - Or allow to extract the secrets who would be normally provided to the plugins who's entrypoint are overwritten. ### Patches https://github.com/woodpecker-ci/woodpecker/pull/3909 https://github.com/woodpecker-ci/woodpecker/pull/3934 ### Workarounds _Is there a way for users to fix or remediate the vulnerability without upgrading?_ **Enable the "gated" repo feature and review each change upfront of running** ### References - https://github.com/woodpecker-ci/woodpecker/pull/3909 - https://github.com/woodpecker-ci/woodpecker/pull/3934 - https://github.com/woodpecker-ci/woodpecker-security/issues/10 (info will be published later at https://github.com/woodpecker-ci/woodpecker/issues/3929) - https://github.com/woodpecker-ci/woodpecker/issues/3929 (info will be published later once we got adoptio...

A SSRF vulnerability in WADL service description in versions of Apache CXF before 4.0.5, 3.6.4 and 3.5.9 allows an attacker to perform SSRF style attacks on REST webservices. The attack only applies if a custom stylesheet parameter is configured.

In versions of Apache CXF before 3.6.4 and 4.0.5 (3.5.x and lower versions are not impacted), a CXF HTTP client conduit may prevent HTTPClient instances from being garbage collected and it is possible that memory consumption will continue to increase, eventually causing the application to run out of memory

An improper input validation of the p2c parameter in the Apache CXF JOSE code before 4.0.5, 3.6.4 and 3.5.9 allows an attacker to perform a denial of service attack by specifying a large value for this parameter in a token. 

### Summary `BinaryHttpParser` does not properly validate input values thus giving attackers almost complete control over the HTTP requests constructed from the parsed output. Attackers can abuse several issues individually to perform various injection attacks including HTTP request smuggling, desync attacks, HTTP header injections, request queue poisoning, caching attacks and Server Side Request Forgery (SSRF). Attacker could also combine several issues to create well-formed messages for other text-based protocols which may result in attacks beyond the HTTP protocol. ### Details **Path, Authority, Scheme** The BinaryHttpParser class implements the readRequestHead method which performs most of the relevant parsing of the received request. The data structure prefixes values with a variable length integer value. The algorithm to create a variable length integer value is below: ``` def encode_int(n): if n < 64: base = 0x00 l = 1 elif n in range(64, 16384): ...

### Impact The two gRPC ports 7070 and 7071, are not bound to [localhost](http://localhost/) by default, so when TorchServe is launched, these two interfaces are bound to all interfaces. Customers using PyTorch inference Deep Learning Containers (DLC) through Amazon SageMaker and EKS are not affected. ### Patches This issue in TorchServe has been fixed in [#3083](https://github.com/pytorch/serve/pull/3083). TorchServe release 0.11.0 includes the fix to address this vulnerability. ### References * [#3083](https://github.com/pytorch/serve/pull/3083) * [TorchServe release v0.11.0](https://github.com/pytorch/serve/releases/tag/v0.11.0) Thank Kroll Cyber Risk for for responsibly disclosing this issue. If you have any questions or comments about this advisory, we ask that you contact AWS Security via our [vulnerability reporting page](https://aws.amazon.com/security/vulnerability-reporting) or directly via email to [[email protected]](mailto:[email protected]). Please do not...

### Impact TorchServe's check on allowed_urls configuration can be by-passed if the URL contains characters such as ".." but it does not prevent the model from being downloaded into the model store. Once a file is downloaded, it can be referenced without providing a URL the second time, which effectively bypasses the allowed_urls security check. Customers using PyTorch inference Deep Learning Containers (DLC) through Amazon SageMaker and EKS are not affected. ### Patches This issue in TorchServe has been fixed by validating the URL without characters such as ".." before downloading: [#3082](https://github.com/pytorch/serve/pull/3082). TorchServe release 0.11.0 includes the fix to address this vulnerability. ### References * [#3082](https://github.com/pytorch/serve/pull/3082) * [TorchServe release v0.11.0](https://github.com/pytorch/serve/releases/tag/v0.11.0) Thank Kroll Cyber Risk for for responsibly disclosing this issue. If you have any questions or comments about this advisory...

A security issue was discovered in Kubernetes clusters with Windows nodes where BUILTIN\Users may be able to read container logs and NT AUTHORITY\Authenticated Users may be able to modify container logs.