CVE-2026-62239

FlashAttention through 2.8.3.post1, fixed in commit 0816ef1, contains a symlink attack vulnerability in the downloadandcopy function within hopper/setup.py that extracts NVIDIA toolchain archives without validating symlinks or filtering tar members. A local attacker can pre-plant a symlink in the...

CVE-2026-62239

FlashAttention (up to 2.8.3.post1) contains a symlink attack in hopper/setup.py: download_and_copy() extracts NVIDIA toolchain archives without validating symlinks or filtering tar members. A local attacker can plant a predictable symlink in the cache to redirect extracted binaries to a chosen lo...

OpenEXR: OpenEXR: Arbitrary code execution via integer overflow in image resizing

A flaw was found in OpenEXR, an image storage format library for the motion picture industry. An integer overflow vulnerability exists in the ImageChannel::resize function, which can be triggered when processing a specially crafted OpenEXR image file through the OpenEXRUtil public API. This can...

OpenEXR: OpenEXR: Arbitrary code execution via integer overflow in image resizing

A flaw was found in OpenEXR, an image storage format library for the motion picture industry. An integer overflow vulnerability exists in the ImageChannel::resize function, which can be triggered when processing a specially crafted OpenEXR image file through the OpenEXRUtil public API. This can...

CVE-2026-49970

Laravel-Mediable before 7.0.0 contains a path traversal vulnerability in the File::sanitizePath function that allows attackers to write uploaded files to arbitrary locations by controlling the directory argument passed to MediaUploader::toDestination. Attackers can exploit the permissive...

CVE-2026-49970

Vulnerability summary: Laravel-Mediable before 7.0.0 contains a path traversal flaw in File::sanitizePath() that lets attackers write uploaded files to arbitrary locations via MediaUploader::toDestination(), due to a permissive regex allowing dot/slash and a weak trailing trim. This can enable re...

PYSEC-2026-2969 psd-tools vulnerable to arbitrary file write via smart-object filename

psd-tools: arbitrary file write/read via smart-object path traversal Summary In psd-tools all releases exposing the SmartObject API through v1.17.0, SmartObject.save writes an embedded smart object to a path taken verbatim from the PSD file. Because that name is attacker-controlled and unsanitise...

PYSEC-2026-3061 Serena: Unauthenticated Flask dashboard on fixed port enables DNS rebinding → memory poisoning → RCE

Summary Serena's built-in web dashboard exposes an unauthenticated Flask API on a fixed, predictable port TCP 24282, hardcoded as 0x5EDA in constants.py. The server has no authentication, no CSRF protection, and no Host header validation. A DNS rebinding attack allows a malicious webpage to reach...

PYSEC-2026-2596 Linuxfabrik Monitoring Plugins allow insecure creation of SQLite databases

Summary The SQLite databases are created at predictable static paths in /tmp. Any user can therefore create a symlink at these paths in /tmp pointing to arbitrary files. The monitoring scripts then follows these symlinks and then creates their database at the symlink target. This becomes really...

PYSEC-2026-2578 Langroid: Path traversal in the file tools allows read/write outside configured current directory

Summary Langroid's ReadFileTool and WriteFileTool appear to treat currdir as the intended working-directory boundary for file operations. However, the tools only change the process working directory to currdir and then operate on the user-supplied filepath without resolving and enforcing that the...

PYSEC-2026-3049 Recce server has unauthenticated SQL execution that allows local file read/write through DuckDB

Impact Recce OSS server deployments that expose the server to an untrusted network without authentication are vulnerable to unauthenticated SQL execution through the query run API. When Recce is configured with a DuckDB-backed project, an attacker can use DuckDB filesystem primitives to read and...

PYSEC-2026-2775 Open Babel has out-of-bounds write in MOPAC translationVectors[] (FINAL POINT)

Summary A memory-safety vulnerability in Open Babel's MOPAC output parser allowed an out-of-bounds write into the translationVectors array when reading the "FINAL POINT" block of a crafted input file. Details The MOPAC output reader stored translation vectors from the FINAL POINT block into a...

PYSEC-2026-2773 Open Babel has out-of-bounds write in ORCA nAtoms parser (second variant)

Summary A memory-safety vulnerability in Open Babel's ORCA parser allowed an out-of-bounds write when reading a crafted input file. Details A second variant of the nAtoms out-of-bounds write in the ORCA reader: a different malformed-input path produced the same class of write past the end of the...

PYSEC-2026-2792 Open Babel has out-of-bounds write in MOL2 attribute/value parser

Summary A memory-safety vulnerability in Open Babel's MOL2 parser allowed an out-of-bounds write when reading a crafted input file. Details The flaw was in the attribute/value parsing path of the MOL2 reader. An over-long attribute or value caused the parser to write past the end of a fixed-size...

PYSEC-2026-2624 mcp-memory-service: Missing Authentication on Document API Endpoints Allows Unauthenticated Memory Read/Write/Delete

Missing Authentication on Document API Endpoints Allows Unauthenticated Memory Read/Write/Delete Summary All HTTP routes under /api/documents/ in mcp-memory-service are served without any authentication dependency, even when the server is configured with an API key MCPAPIKEY or OAuth. An...

PYSEC-2026-2788 Open Babel has out-of-bounds write in CSR PadString (title field)

Summary A memory-safety vulnerability in Open Babel's CSR parser allowed an out-of-bounds write when reading a crafted input file. Details The flaw was in the PadString helper used to handle the CSR title field. A title longer than the fixed destination buffer caused the parser to write past the...

PYSEC-2026-2622 mcp-memory-service: OAuth read-only clients can write and delete memories through MCP tools/call

Summary The HTTP MCP JSON-RPC endpoint at /mcp requires only OAuth read scope for all requests, then dispatches tools/call directly to handlers that include mutating tools. A read-only OAuth client can call storememory and deletememory through MCP even though the corresponding REST endpoints...

PYSEC-2026-2974 py7zr: Arbitrary File Write Vulnerability

Summary There exists an arbitrary file write vulnerability in py7zr 1.1.0, latest, which allows symbolic links to be recreated outside the destination directory via crafted malicious symbolic link chains. When using extractall to extract an archive, the library restores these symbolic links,...

PYSEC-2026-2393 BBOT: Arbitrary File Write in postman_download Module

The postmandownload module uses the workspace name field from the Postman API to construct the local directory path without sanitization. If a malicious workspace has a name containing path traversal characters, pathlib resolves the path outside the intended output directory, allowing an attacker...

PYSEC-2026-2394 BBOT: Symlink-Following Arbitrary Write via github_workflows Module

The githubworkflows module constructs local directory paths from user-controlled repository names without validating for symlinks. A local attacker sharing the scan directory can plant a symlink at the predictable output path, causing workflow data to be written to an attacker-chosen location...