Sensitive information written to logs by Snowflake drivers
Insertion of sensitive information into log files in the Snowflake Python, Go, JDBC, Node.js, PHP PDO, and ODBC drivers allowed authentication tokens, query-result encryption keys, pre-signed cloud-storage URLs, and SAML assertions to be written to diagnostic logs in circumstances where the available log redaction did not cover all affected log paths and data types. An attacker with read access to the log destination, whether the local filesystem, a log aggregation service, or a CI/CD artifact store, could obtain credentials and decryption keys that, if still valid at the time of access, could be used to authenticate to the corresponding Snow
wger before 2.6 CSV Formula Injection via member export
wger before 2.6 fails to sanitize first_name and last_name fields in the gym member TSV export endpoint, allowing any gym member to inject spreadsheet formulas. Attackers can inject formulas like =HYPERLINK to exfiltrate admin data or execute code when admins open the exported file in Excel or LibreOffice Calc.
wger before 2.6 Open Redirect via trainer-login next parameter
wger before 2.6 (affected versions <= 2.5.0) contains an open redirect vulnerability in the trainer_login view (wger/core/views/user.py). After a trainer enters impersonation mode, the view redirects to the user-supplied 'next' GET parameter via HttpResponseRedirect() without validating it with url_has_allowed_host_and_scheme(). An attacker who delivers a crafted link to an authenticated trainer can redirect the trainer's browser to an attacker-controlled domain, enabling phishing and leaking the wger URL structure (including the impersonated user's user_pk) via the Referer header.
wger before 2.5 Uncontrolled Resource Consumption via date_sequence
wger before 2.5 fails to validate the maximum duration of routine date ranges, allowing authenticated users to create routines spanning arbitrarily long periods. Attackers can trigger the date_sequence computation via routine detail endpoints, forcing the server to iterate thousands of times per request and exhaust worker threads, denying service to legitimate users.
Arbitrary file read via workspace confinement bypass in local-operator /v1/chat/agents/{id}/edit
The `/v1/chat/agents/{agent_id}/edit` endpoint in local-operator versions before 0.47.5 resolves the caller-supplied `file_path` with `expanduser().resolve()` and reads it without checking that it lies inside the agent's workspace. An unauthenticated client with network access to the API can supply an absolute path or a path containing traversal sequences and obtain the contents of any file readable by the server process, which are placed in the model prompt and returned in the response.
Version 0.47.5 resolves server-side reads within the agent's configured workspace and rejects canonical paths outside it (including symlink and junction esc
Path traversal and arbitrary directory deletion/overwrite via agent profile import in local-operator
The `/v1/agents/import` endpoint and `AgentRegistry.import_agent()` in local-operator versions before 0.47.5 trust the `id` field inside `agent.yml` of an uploaded agent profile archive when constructing the destination directory. A crafted `id` containing directory traversal sequences makes `shutil.rmtree` and `shutil.copy2` operate outside the agent registry, allowing an unauthenticated client with network access to the API to recursively delete arbitrary directories and write files with the privileges of the server process. Even without traversal, an imported archive could overwrite or delete existing local agent profiles.
Version 0.47.5
pymonocypher: Potential heap buffer overflow on nb_blocks in argon2i_32 when provided buffer is too small
pymonocypher uses cython to wrap the Monocypher C library. Prior to version 4.0.2.8, the argon2i_32 implementation does not check the nb_blocks size. If the caller does not provide a sufficiently large buffer based on the API contract, then argon2i_32 will write past the end of the buffer and possibly corrupt the heap. This issue has been patched in version 4.0.2.8.
Windows ML CLI: CORS misconfig enables localhost RCE
Windows ML CLI is a command line tool for building portable, performant, and high-quality AI models for Windows ML. Prior to 0.4.0, the src/winml/modelkit/serve/cli_api.py component exposes WinML CLI commands through a localhost HTTP API without authentication and configures the allow_origins setting as a wildcard in both src/winml/modelkit/serve/cli_api.py and src/winml/modelkit/serve/app.py. A malicious website loaded by a user can send cross-origin requests to /v1/cli/build or /v1/cli/config and set the trust_remote_code parameter to true, which is converted to the --trust-remote-code command-line flag without validation. This reaches Auto
HTTPX2: Streaming response decompression does not bound peak memory (decompression amplification)
HTTPX2 is a next generation HTTP client for Python. Prior to 2.12.0, the HTTPX2 content decoders in src/httpx2/httpx2/_decoders.py fully inflate each gzip, deflate, br, or zstd network chunk before iter_bytes() or aiter_bytes() yields bounded pieces to the application. A 64 KiB compressed chunk can expand to approximately 64 MiB in one intermediate allocation, so an attacker-controlled or compromised server can cause severe memory pressure or out-of-memory process termination even when the application streams the response. This issue is fixed in version 2.12.0.
HTTPX2: Secure WebSocket traffic sent without TLS through SOCKS proxies
HTTPX2 is a next generation HTTP client for Python. Prior to 2.10.0, httpcore2 fails to start TLS in src/httpcore2/httpcore2/_sync/socks_proxy.py and src/httpcore2/httpcore2/_async/socks_proxy.py when the remote origin uses wss through a SOCKS5 proxy because the TLS upgrade condition only recognizes https. HTTPX2 exposes the flaw through Client.websocket() and AsyncClient.websocket() from 2.6.0 through 2.9.1, so the opening handshake, query parameters, Authorization headers, cookies, and subsequent frames can cross the proxy path in plaintext without certificate verification. An attacker controlling or observing that path can read or modify t
HTTPX2: Conflicting Content-Length and Transfer-Encoding headers can be auto-generated
HTTPX2 is a next generation HTTP client for Python. Prior to 2.11.0, Request._prepare() in src/httpx2/httpx2/_models.py can add a body-derived Content-Length header to a request that already contains a caller-supplied Transfer-Encoding header because its setdefault() processing checks each default header independently rather than treating the two framing headers as mutually exclusive. Fixed-size byte, JSON, form, and known-length multipart bodies can therefore be serialized over HTTP/1.1 with both headers, allowing request smuggling or connection desynchronization when downstream intermediaries disagree about which framing header takes preced
HTTPX2: Multipart part header injection via unvalidated file Content-Type and custom headers
HTTPX2 is a next generation HTTP client for Python. Prior to 2.11.0, FileField.render_headers() in src/httpx2/httpx2/_multipart.py directly interpolates attacker-controlled content_type values and custom headers from the files= three-element (filename, content, content_type) tuple and the files= four-element (filename, content, content_type, headers) tuple into multipart/form-data part headers without validating header names or values. CR or LF characters can terminate a part header, inject additional part headers, or end the part header block early, allowing a downstream multipart parser to treat attacker-supplied lines as genuine headers an
HTTPX2: Quadratic SSE line buffering can cause CPU denial of service
HTTPX2 is a next generation HTTP client for Python. From 2.5.0 until 2.10.0, the HTTPX2 Server-Sent Events parser in src/httpx2/httpx2/_sse.py repeatedly copies and rescans buffered text in _SSELineDecoder.decode() when an attacker-controlled or compromised SSE endpoint splits one unterminated line across many response chunks. The behavior affects httpx2.Client.sse() and httpx2.AsyncClient.sse(), and the total processing work grows quadratically with the line length, allowing a crafted stream to consume excessive CPU and block a synchronous worker or asynchronous event loop. This issue is fixed in version 2.10.0.
LiteLLM: Authenticated SSRF and provider-credential exfiltration via unvalidated request-body routing parameters
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to versions 1.88.6 and 1.96.2, any authenticated LiteLLM proxy user could redirect an outbound provider call to a destination the user controls and cause the proxy to send its configured provider credentials to that destination. Request validation in litellm/proxy/auth/auth_utils.py, litellm/proxy/common_request_processing.py, litellm/proxy/health_endpoints/_health_endpoints.py, litellm/proxy/image_endpoints/endpoints.py, and litellm/proxy/litellm_pre_call_utils.py used incomplete checks that did not cover every sensitive parameter or inspect equivalen
Apache Spark: XSS Vulnerability in Spark Web 3.5.4
There is a lack of XSS escaping in the Spark History Server prior to 3.5.8 which allows a malicious Spark job to generate arbitrary unescaped frontend code which could lead to a minimal privilege escalation in browser. Users are encouraged to upgrade to Spark 3.5.8 or later.
This CVE is marked as "low" since the path to exploit requires both relatively high permissions (ability to launch a Spark job) and requires tricking a user with higher permissions to log in and visit the Spark history web page.
Users are encouraged to upgrade their Spark history servers to Spark 3.5.8 or later.
Scrapy: S3DownloadHandler sends signed S3 requests over plaintext HTTP by default
Scrapy is a high-level web crawling and scraping framework for Python. Prior to 2.17.0, in scrapy/core/downloader/handlers/s3.py, Scrapy's S3DownloadHandler converts an S3-scheme bucket and key request into a plaintext HTTP request to the corresponding S3 endpoint unless request.meta["is_secure"] is explicitly enabled, then signs and sends the plaintext request with configured AWS credentials. A network attacker who can observe traffic between Scrapy and S3 can read the bucket and key path, AWS Authorization header, X-Amz-Security-Token when temporary credentials are used, S3 object contents, and S3 response headers. An active man-in-the-midd
pypdf: Possible long runtimes/large memory usage when extracting XForm objects
pypdf is a free and open-source pure-python PDF library. Prior to 6.16.1, an attacker can craft a PDF that causes pypdf/_page.py PageObject._extract_text and PageObject.extract_xform_text to traverse a directed acyclic graph of reused form XObjects in which each form invokes a child multiple times, creating exponentially many traversal paths and causing long runtimes and large memory consumption. This issue is fixed in version 6.16.1.
pypdf: Possible long runtimes/large memory usage when retrieving outlines
pypdf is a free and open-source pure-python PDF library. Prior to 6.16.1, an attacker can craft a PDF that causes pypdf/_doc_common.py _get_outline to consume long runtimes and large amounts of memory when retrieving document outlines with large numbers of entries or deeply nested reused paths because the traversal lacked global entry-count and nesting-depth limits. This issue is fixed in version 6.16.1.
pypdf: Possible infinite loop for TreeObject.insert_child
pypdf is a free and open-source pure-python PDF library. Prior to 6.16.0, an attacker can craft a PDF whose cyclic tree structure causes pypdf/generic/_data_structures.py TreeObject.insert_child to follow /Next links indefinitely when a writing code path inserts a child, producing an infinite loop. This issue is fixed in version 6.16.0.
sqlparse: Reindentation of tuple lists causes near-cap quadratic CPU consumption
sqlparse is a non-validating SQL parser module for Python. Prior to 0.6.0, sqlparse.format(sql, reindent=True) and sqlformat --reindent route attacker-controlled parenthesized tuple lists through ReindentFilter._get_offset() in sqlparse/filters/reindent.py, where _flatten_up_to_token() repeatedly rebuilds and joins the statement prefix. Thousands of offset calculations walk an expanding token tree, producing quadratic CPU consumption for inputs that remain below MAX_GROUPING_TOKENS and causing request delays, reduced throughput, or worker starvation. This issue is fixed in version 0.6.0.