Potential exposure of private data via cached Set-Cookie response
An issue was discovered in Django 6.0 before 6.0.7 and 5.2 before 5.2.16.
`UpdateCacheMiddleware` and the `cache_page()` decorator cache responses that vary on cookies when the incoming request carries unrelated cookies, which allows remote attackers to read private data from the shared cache.
Earlier, unsupported Django series (such as 5.0.x, 4.1.x, and 3.2.x) were not evaluated and may also be affected.
Django would like to thank Chris Whyland for reporting this issue.
Kinto Attachment's attachments can be replaced on read-only records
### Impact
The attachment file of an existing record can be replaced if the user has `"read"` permission on one of the parent (collection or bucket).
And if the `"read"` permission is given to `"system.Everyone"` on one of the parent, then the attachment can be replaced on a record using an anonymous request.
Note that if the parent has no explicit read permission, then the records attachments are safe.
### Patches
- Patch released in kinto-attachment 6.4.0
- https://github.com/Kinto/kinto-attachment/commit/f4a31484f5925cbc02b59ebd37554538ab826ca1
### Workarounds
None if the read permission has to remain granted.
Updating to 6.4.0 or
jwcrypto token substitution can lead to authentication bypass
The JWT code can auto-detect the type of token being provided, and this can lead the application to incorrect conclusions about the trustworthiness of the token.
Quoting the private disclosure we received : "Under certain circumstances, it is possible to substitute a [..] signed JWS with a JWE that is encrypted with the public key that is normally used for signature validation."
This substitution attack can occur only if the validating application also have access to the private key, normally used to sign the tokens, available during validation of the received JWT.
The significance of this attacks depends on the use of the token, it may lead
Apache Airflow: DAG author RCE on webserver via unrestricted import_string() in BaseSerialization.deserialize()
A bug in `BaseSerialization.deserialize()` allowed unrestricted `import_string()` of attacker-controlled class paths when the Scheduler / API Server loaded a serialized DAG: a DAG author could embed a malicious trigger into a DAG to gain remote code execution on the API Server / Scheduler process, crossing the Airflow security boundary that DAG-author code must never execute in those processes. Users are advised to upgrade to `apache-airflow` 3.3.0 or later. As a defense-in-depth mitigation, deployments where DAG-author trust is limited can restrict the `[core] allowed_deserialization_classes` config to a narrow allowlist.
Apache Airflow: Task-instance API exposes secrets in deferred trigger kwargs
In Apache Airflow before 3.3.0, the REST API task-instance detail and list
endpoints returned a deferred task's trigger kwargs without masking. When a
deferred operator passed a secret (for example a provider API key) into its
trigger, any authenticated user with DAG-scoped task-instance read access for
that DAG could read that secret in clear text while the task was deferred.
Users should upgrade to apache-airflow 3.3.0 or later, which masks sensitive
values in trigger kwargs returned by the API.
Apache Airflow: Bulk JSON Variables bypass should_hide_value_for_key - redact() called without the key
The Bulk Variables API in Apache Airflow called the redactor without passing the variable's key, so the key-based `should_hide_value_for_key` check (which triggers on secret-suffixed key names like `*_password` / `*_token` / `*_secret`) could not fire for JSON-decodable variable values. An authenticated UI/API user with bulk Variable read permission could retrieve plaintext values from JSON variables whose key would otherwise trigger redaction. Affects deployments that store sensitive values in JSON-typed Airflow Variables under secret-suffixed key names. Users are advised to upgrade to `apache-airflow` 3.3.0 or later (the fix landed on `main
Apache Airflow: Per-DAG read bypass discloses co-located DAGs' source via GET /api/v2/dagSources/{dag_id}
Before apache-airflow 3.3.0, a user authorized to read one Dag could disclose the source of other Dags co-located in the same source file. `GET /api/v2/dagSources/{dag_id}` — and the equivalent Dag-source view in the UI — returned the entire source file without redacting Dags the caller was not authorized to read, bypassing per-DAG read authorization. Deployments that co-locate multiple Dags in a single file and rely on per-DAG access control to limit source visibility are affected; single-Dag-per-file deployments are not. Upgrade to apache-airflow 3.3.0 or later.
Apache Airflow: /ui/dependencies scheduling graph leaks unreadable Dag identifiers via trigger/sensor dep.source/dep.target
A bug in Apache Airflow's `/ui/dependencies` scheduling graph endpoint applied the caller's readable-Dag filter to the top-level serialized Dag key but still emitted referenced Dag IDs through the `dep.source` and `dep.target` fields of trigger / sensor dependency entries. An authenticated UI user with read permission on some Dags could enumerate the identifiers of other Dags they were not authorized to read by inspecting the dependency graph for trigger / sensor references. Affects deployments that rely on per-Dag read scoping to keep Dag identifiers private across teams. This is a residual gap in the fix for CVE-2026-28563, which filtered t
Apache Airflow: Config API leaks per-key secrets backend kwargs - masker bypass on synthetic options
The Config API in Apache Airflow surfaced per-key secrets-backend overrides (environment variables like `AIRFLOW__SECRETS__BACKEND_KWARG__SECRET_ID` and `AIRFLOW__WORKERS__SECRETS_BACKEND_KWARG__SECRET_ID`) as synthetic config options whose option names were not in `sensitive_config_values`, so the masker did not redact them. An authenticated UI/API user with Config read permission could retrieve plaintext secrets-backend credentials (Vault `role_id` / `secret_id`, etc.) from the Config API output. Affects deployments that configure secrets backends via per-key environment overrides. Users are advised to upgrade to `apache-airflow` 3.3.0 or l
Crawl4AI: Unauthenticated RCE via Chromium launch-argument injection in browser_config.extra_args
Crawl4AI is an open-source LLM-friendly web crawler and scraper. Prior to 0.9.0, the Docker API server accepted request-supplied browser_config.extra_args, which flowed into Chromium's launch arguments. An attacker could inject Chromium switches that replace a child-process launch command together with --no-zygote, causing Chromium to fork or exec an attacker-controlled command as the container's runtime user. The Docker API is unauthenticated by default, so a single request yields arbitrary command execution. This issue is fixed in version 0.9.0.
Crawl4AI unauthenticated SSRF in Docker streaming crawl endpoint
Crawl4AI is an open-source LLM-friendly web crawler and scraper. Prior to 0.9.0, the Docker API server applied its SSRF destination check on the non-streaming /crawl path but not on the streaming path. handle_stream_crawl_request passed seed URLs straight to the crawler with no destination validation, allowing a remote unauthenticated client to call POST /crawl/stream or POST /crawl with crawler_config.stream=true with a URL pointing at an internal, private, or link-local address; the server fetched it and streamed the response body back. This issue is fixed in version 0.9.0.
Crawl4AI arbitrary file write via download filename path traversal
Crawl4AI is an open-source LLM-friendly web crawler and scraper. Prior to 0.9.0, when the crawler saves a downloaded file, the destination filename was taken from attacker-influenced input and joined to the downloads directory with no confinement. A filename containing an absolute path or traversal escaped the downloads directory, giving an arbitrary file write with attacker-controlled contents; the HTTP crawler path uses the response Content-Disposition filename and the browser crawler path uses the download's suggested filename. Because the written bytes are attacker-controlled, this can escalate to remote code execution. This issue is fixe
vLLM denial of service via prompt embeds on M-RoPE models
vLLM is a library for LLM inference and serving. From 0.12.0 to before 0.24.0, sending a pure prompt embeds payload in a /v1/completions request with a model using M-RoPE causes EngineCore to fail an assertion and fatally crash, shutting down the entire server application. Any remote user who is authorized to make a /v1/completions request can make such a request and induce a crash. This issue is fixed in version 0.24.0.
vLLM: ReDoS via structured_outputs.regex compiled without timeout in xgrammar and outlines backends
vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Prior to 0.24.0, the structured_outputs.regex API parameter passes a user-supplied regular expression string directly to the grammar compiler backends with no compilation timeout; in the xgrammar backend the string reaches the regex compiler with no guard, and in the outlines backend the validation step blocks structural issues such as lookarounds and backreferences but performs no complexity analysis, so a pattern with nested quantifiers passes all checks and causes exponential state-space expansion, allowing a single request containing an adversarial regex
vLLM: Remote DoS in vLLM via Invalid Recovered Token Reinjection
vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Prior to 0.24.0, a frontend-legal multi-request speculative decoding workload can cause the rejection sampler to produce a recovered token equal to the model vocabulary size boundary value, which is then converted to negative one when the engine selects the next live token for a request and is written back into the drafter's input ids; that out-of-vocabulary value is later consumed by the model's embedding and attention path and crashes the engine worker with a GPU device-side assertion. The same triggering request sequence is reachable through the public gR
vLLM speech-to-text endpoints allocate full upload before enforcing the audio file-size limit
vLLM is an inference and serving engine for large language models. From 0.22.0 to 0.23.0, the /v1/audio/transcriptions and /v1/audio/translations routes call request.file.read() to fully materialize an uploaded audio file into memory before vLLM checks the documented VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed upload size limit (default 25 MB) later in the speech-to-text preprocessing step, so an API caller who can reach those routes can submit an oversized multipart upload and cause vLLM to allocate memory proportional to the uploaded file size before the request is rejected as too large, creating memory pressure or terminating the process de
Pillow BdfFontFile`: `Image.new()` called without `_decompression_bomb_check()` — bomb protection bypass via font loading
Pillow is a Python imaging library. Prior to 12.3.0, PIL/BdfFontFile.py bdf_char() read the BBX width and height field from a BDF font file and passed attacker-controlled dimensions to Image.new() without calling Image._decompression_bomb_check(), bypassing Pillow's documented decompression bomb protection and allowing excessive memory allocation. This issue is fixed in version 12.3.0.
Pillow is a Python imaging library. Prior to 12.3.0, PIL/GdImageFile.py GdImageFile._open() read image dimensions from the GD 2.x header and stored them in self._size without calling Image._decompression_bomb_check(), allowing a crafted .gd file to trigger excessive C-heap allocation when loaded. This issue is fixed in version 12.3.0.
Pillow: `FontFile.compile()`: `Image.new()` called without `_decompression_bomb_check()`
Pillow is a Python imaging library. Prior to 12.3.0, PIL/FontFile.py FontFile.compile() assembled per-glyph images into a combined bitmap with Image.new("1", (xsize, ysize)) without calling Image._decompression_bomb_check(), allowing a font to trigger excessive allocation during conversion or saving. This issue is fixed in version 12.3.0.