| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| TAKO AI Agent for Okta is a local-first AI agent that helps Okta administrators query tenant data using natural language. Prior to 3.0.1-beta, any authenticated user with a valid `fctr_session` cookie can submit arbitrary Python source in the `script_code` field of `POST /api/react/execute-script`, without administrator privilege. In `src/api/routers/react_stream.py`, the source is stored with `skip_discovery=True`, bypassing AI-based content filtering, and is passed to `validate_generated_code` in `src/utils/security_config.py`. `BLOCKED_PATTERNS` recognizes direct `__import__` calls but misses the import name passed as a string, `getattr` is in `ALLOWED_BUILTINS`, and the AST checker handles `ast.Name` and `ast.Attribute` targets but fails to handle `ast.Call` targets. Dynamic `getattr` call chains through `__builtins__` therefore evade validation. The `_write_temp_script` helper writes the accepted code to disk, and `asyncio.create_subprocess_exec()` executes it as the application process user, which is root in the default container. Command output is returned as Server-Sent Events through `GET /api/react/stream-react-updates?process_id=<id>`. The resulting container compromise can disclose environment secrets including Okta API tokens and the JWT key, expose the SQLite database, and permit backend-service pivoting and persistence. This issue is fixed in version 3.0.1-beta. |
| OpenSearch SQL supports querying OpenSearch data using SQL and Piped Processing Language. In the 2.x line from 2.8.0.0 until 2.19.6.0, and the 3.x line prior to 3.7.0.0, a remote authenticated user with basic read/search permissions can provide a crafted `cursor` parameter to the `plugins/sql` endpoint and trigger unsafe deserialization of untrusted data, resulting in arbitrary code execution on the server. This issue is fixed in repository releases 2.19.6.0 and 3.7.0.0. |
| Coop is an open source review and moderation tool from ROOST. Prior to 1.0.2, `addAccessibleQueuesForUser` and `removeAccessibleQueuesForUser` accepted arbitrary `queueId` and `userId` values without organization scoping. An authenticated user belonging to one organization and holding MRT queue-edit permission could grant or revoke another organization's users' access to manual review queues. The existing queue-edit permission check did not bind the supplied identifiers to the caller's organization. The attacker could not read the other organization's queues or jobs, so this issue did not disclose confidential data. This issue is fixed in version 1.0.2. |
| Coop is an open source review and moderation tool from ROOST. Prior to 1.0.2, the SAML strategy's verify callback resolved the authenticated user by email alone using `kyselyUserFindByEmail`, without binding the lookup to the organization in `/saml/login/:orgId/callback`. An attacker who administers an organization and controls its SAML IdP signing key can submit an assertion for a victim email to another organization's callback when that target organization has registered the same signing certificate. The assertion verifies against the target certificate, and the email resolves to the target user, creating a session as that user and allowing full cross-tenant account takeover without privileges in the target organization. Exploitation requires the unusual configuration in which the target certificate matches a signing key controlled by the attacker. This issue is fixed in version 1.0.2. |
| Coop is an open source review and moderation tool from ROOST. Prior to 1.0.2, `resetPasswordForToken` updates the password row without invalidating database-backed sessions. An attacker who has acquired a password and established a session retains access even after the legitimate user resets the password, leaving the compromised account under attacker control until the session expires at its `30-day` maximum age. This issue is fixed in version 1.0.2. |
| secp256k1-native is a native C secp256k1 implementation for Ruby. In affected native builds, native constant-time scalar multiplication through `Point#mul` or `scalar_multiply_ct` can leak timing information about a secret scalar when compiled by a toolchain that reconstructs branchless selection into a branch. The `Montgomery ladder` uses `uint256_select` with an all-zeros or all-ones mask, but affected GCC builds at `-O2` recover the original condition and emit secret-dependent `je` or `jne` jumps whose timing correlates with scalar bits. This defeats the assumption behind the 0.17.0 `jp_add_internal` infinity-branch fix. A consumer must multiply a secret private key or nonce in a context where an attacker can observe timing; repeated measurements can statistically recover scalar bits. The defect is in the compiled binary, so source review or `ctgrind` on a non-reconstructing toolchain may miss it. GCC 14.3.0, 15.1.0, and 15.2.0 reproduced the branch reconstruction, and `extconf.rb` appends `-O2`. Tested GCC versions through 12.4.0 used bitwise operations and GCC 13.3.0 used constant-time `cmov`; the pure-Ruby path is not affected by this flaw and separately makes no constant-time claim. GCC 16 and clang were untested, and point releases, targets, and flag combinations were not exhaustively bisected. `Earlier pre-0.17.0 versions had different constant-time issues rather than this introduced select idiom.` Version 0.18.0 contains the verified fix. The advisory's broad pre-0.18.0 guidance also covers earlier, different constant-time issues; whether this CVE groups those issues or only the select idiom introduced in 0.17.0 remains unresolved. |
| Improper Authorization vulnerability in Drupal Field Collection for Drupal 7 allows a user with the "edit field collections" permission to bypass access control on the parent (host) entity.
The access check for field collection items (field_collection_item_access()) granted edit and delete access to any user holding the "edit field collections" permission without verifying the user's access to the host entity. As a result, a user with "edit field collections" but without edit/update access to the host entity (for example a node they cannot edit) could edit or delete field collection items attached to that entity, bypassing the intended entity/node access restrictions. The fix requires access to the parent entity in addition to the field collection permission.
This issue affects Field Collection for Drupal 7: version 7.x-1.0 up to and including 7.x-1.2. |
| The File Force Download module for Drupal 7 implements hook_file_download() to add a Content-Disposition header when a file URL carries the download query parameter. In Drupal 7, returning a non-empty header array from that hook is itself a grant of access to the file.
In versions 7.1.0 through 7.1.2, the module only delegated to other modules' hook_file_download() implementations when the file was in the private scheme. For every other scheme it returned its headers without any check. An unauthenticated visitor who appended download=1 to the URL of a file in the temporary:// scheme, or in a stream wrapper supplied by another module, could retrieve the file, overriding the access checks that would otherwise have applied. Files in the public and private schemes are not affected. |
| The ThirstyAffiliates – Affiliate Links, Link Branding, Link Tracking & Marketing Plugin plugin for WordPress is vulnerable to Stored Cross-Site Scripting via the 'comment' parameter in all versions up to, and including, 3.12.1 due to insufficient input sanitization and output escaping. This makes it possible for unauthenticated attackers to inject arbitrary web scripts in pages that will execute whenever a user accesses an injected page. This vulnerability is only exploitable on block themes (such as Twenty Twenty-Five) where the Comment Content block causes comment HTML to flow through render_block, triggering the plugin's transformer. Additionally, a successful attack requires the attacker to have had one prior comment approved, after which subsequent crafted comments are auto-approved and rendered publicly. |
| Heap-based buffer overflow in Opengear IM72XX, CM71XX, and ACM70XX firmware versions 4.8.0 through 5.5.2 on console server appliances allows an unauthenticated remote attacker to crash the index.cgi process via a crafted multipart/form-data POST request containing an overlong Content-Disposition name field. |
| Nexterm through 1.2.2-BETA contains a missing authorization vulnerability that allows authenticated users to hibernate, resume, or terminate other users' remote sessions via session lifecycle routes. Attackers who obtain a session ID can call the hibernate, resume, or DELETE connection endpoints to disrupt live SSH, RDP, VNC, or SFTP sessions. |
| DeepInfer through 0.1.0 contains a path traversal vulnerability in the unauthenticated /profile endpoint that allows remote attackers to create arbitrary directories via the output_dir parameter. Attackers, including cross-site GET requests from an operator's browser, can write chrome trace files to exhaust disk or inodes and use error messages as a filesystem oracle. |
| Thelia Page Module before 2.0.4 contains a cross-site request forgery vulnerability that allows remote attackers to delete page tags, documents and images by abusing back-office delete routes lacking CSRF tokens and permission checks. Attackers can lure a logged-in administrator into navigating to crafted GET requests that walk sequential ids, removing TheliaLibrary images, their files and variants across pages, products and categories. |
| Improper certificate validation in Visual Studio allows an unauthorized attacker to perform spoofing over a network. |
| XS::libbrotli versions before 1.2.1.0 for Perl bundle a brotli library with an out-of-bounds write in the decoder.
Affected versions compile a bundled brotli into static archives that dependent XS modules link against at build time, rather than using the system library. The bundled copy, 1.0.7 or 1.1.0 across the range, carries the decoder out-of-bounds write described in CVE-2026-111497. Version 1.2.1.0 updates it to brotli 1.2.1.
XS::libbrotli has no Perl interface, so the decoder is reached through any dependent module that decompresses untrusted data with it. An attacker supplying a crafted brotli stream can corrupt decoder state, disclose memory, crash the process, or potentially execute arbitrary code. Upgrading the system brotli does not remediate this. |
| Thelia before 3.1.1 contains a stored cross-site scripting vulnerability that allows authenticated back-office administrators with store update permission to inject scripts by uploading unsanitized SVG logo_file or banner_file images. Attackers can upload an SVG with an onload payload served from /admin/store-media/logo or the public cache/images/store/ directory to execute scripts against visitors and higher-privileged administrators. |
| Thelia Back-office Default Twig Template before 1.1.1 contains a cross-site request forgery vulnerability in the /admin/product/combinations/update and /admin/product/default-price/update routes, which accept GET requests without anti-CSRF token checks. Attackers can lure a logged-in administrator into following a crafted link to rewrite product combination prices, stock, weight, EAN and sale flags, zeroing omitted prices. |
| imagecodecs 2026.3.6 through 2026.10.10 contains an infinite loop vulnerability in ccitt_fax4_decode_size() that allows attackers to hang the calling process by supplying a crafted CCITT Group 4 stream. Attackers can provide an 8-byte G4 stream lacking EOFB to ccittfax4_decode() or the numcodecs Ccittfax4 codec, consuming full CPU until the process is killed. |
| imagecodecs 2023.1.23 through 2026.10.10 contains a stack-based buffer overflow vulnerability in the lossless JPEG Define Huffman Table parser in decode_jpegsof3(). Attackers can supply a crafted lossless JPEG whose DHT symbol counts wrap an 8-bit sum, overflowing the 32-entry HufVal and HufSz stack arrays and crashing the calling process. |
| The VobSub subtitle decoder libsub_vob (part of the Amlogic DroidLogic SubTitleService), as shipped on Nexbox A95X (Amlogic S905X) TV boxes, contains an infinite loop vulnerability in get_vob_spu() that allows unauthenticated network attackers or local users loading malicious files to trigger an unreachable loop exit condition via mismatched length fields. Attackers can send crafted AMLU data blocks over the unauthenticated tcp/10100 listener or via a malicious VobSub (.sub) file to cause the subtitle decode thread to spin at 100% CPU indefinitely, resulting in denial of service. |
