| CVE-2026-49360 | Recce server has unauthenticated SQL execution that allows local file read/write through DuckDB | high | |
| CVE-2026-49292 | Kiwi TCMS's /init-db/ page renders and responds to requests after first use | medium | |
| CVE-2026-54617 | LaunchServer FileServerHandler has an unauthenticated path traversal issue | critical | |
| CVE-2026-52792 | Algernon vulnerable to server-side script source disclosure on Windows via NTFS filename | high | |
| CVE-2026-52834 | jxl-grid on 32-bit platforms has an out-of-bounds writes due to integer overflow | high | |
| CVE-2026-49289 | SimpleSAMLphp has Possible DoS via XPath Transform | high | |
| CVE-2026-52829 | Zebra Address Book Aborted by IPv4-Mapped Mempool Misbehavior Update | high | |
| CVE-2026-49283 | SimpleSAMLphp HTTP-Artifact TLS validator confusion allows cross-IdP authentication bypass | high | |
| CVE-2026-52817 | Linuxfabrik Monitoring Plugins: Sudoers may be able to obtain privilege escalation via /usr/bin/apt-get arguments | high | |
| CVE-2026-59102 | Forgejo before 15.0.3 contains a stored cross-site scripting vulnerability that allows authenticated attackers to execute arbitrary JavaScript in other users' browsers by setting a full name containing an HTML payload and triggering an Actions run. When the DEFAULT_SHOW_FULL_NAME option is enabled, the run description is assembled server-side with the user's display name interpolated into an HTML string via a translation function that does not escape its arguments, and the frontend renders the result using a Vue v-html binding, causing script execution for any user who views the affected Actions run page. | low | 2026-07-06 |
| CVE-2026-59101 | AutoBangumi before 3.2.8 contains a server-side request forgery (SSRF) vulnerability that allows unauthenticated remote attackers to probe internal network services by supplying arbitrary host values to an unprotected setup endpoint. Attackers can send requests to the POST /api/v1/setup/test-downloader endpoint during the initial setup window, causing the server to issue HTTP GET requests to internal or reserved addresses and leak information through echoed connection-error messages. | medium | 2026-07-14 |
| CVE-2026-59100 | LobeChat through 2.2.9 contains a broken object level authorization vulnerability that allows authenticated attackers to access and modify other users' chat-group agent data by supplying arbitrary group identifiers. Attackers can invoke the getGroupAgents, updateAgentInGroup, and removeAgentsFromGroup operations without user-scoped predicates to read agent listings, modify agent roles and ordering, and remove agents from chat groups belonging to other users. | low | 2026-07-14 |
| CVE-2026-59099 | Apereo CAS 7.3.0 before 8.0.0-RC6 contains a cryptographic vulnerability that allows remote unauthenticated attackers to recover plaintext conversation state by exploiting AES-GCM initialization vector reuse across the server lifetime. Attackers can collect multiple client-side webflow execution tokens from the unauthenticated login page and perform known-plaintext analysis to decrypt the webflow conversation state due to keystream reuse caused by a fixed all-zero IV paired with the same encryption key. | critical | 2026-07-14 |
| CVE-2026-59098 | LobeChat through 2.2.9 contains a broken access control vulnerability in the retrieval-augmented-generation semantic search functionality that allows authenticated attackers to access other users' data by exploiting missing user-identifier predicates in the chunk model semanticSearch method. Attackers can supply arbitrary victim file or knowledge-base identifiers through the chunk retrieval and chat knowledge-base paths to retrieve text content, file names, and metadata belonging to other users. | high | 2026-07-14 |
| CVE-2026-59097 | Taiga before 6.10.2 contains a missing authorization vulnerability that allows unauthenticated remote attackers to create default due-date records in any project by exploiting unprotected POST endpoints on the user-story, task, and issue due-date API viewsets. Attackers can supply an arbitrary project identifier to these endpoints, which bypass permission checks and apply the AllowAny default, to pre-empt project administrators from initializing due dates by creating records before they can do so themselves. | medium | 2026-07-14 |
| CVE-2026-59096 | Dapr Sentry's OIDC discovery endpoint derives the issuer and jwks_uri of the /.well-known/openid-configuration document from the request Host, honoring an attacker-controlled X-Forwarded-Host header without validation when no allowed-hosts list is configured (the default), and serves the document with a one-hour public cache lifetime. A remote unauthenticated attacker can poison the discovery document so relying parties performing dynamic (unpinned) discovery fetch the JWKS from an attacker-controlled server, causing attacker-signed JWTs to be accepted. Exploitation requires the OIDC server enabled without a configured jwt-issuer or oidc-allowed-hosts. | high | 2026-07-14 |
| CVE-2026-59095 | LobeChat before 2.2.10-canary.18 contains a server-side request forgery vulnerability that allows authenticated attackers to direct internal HTTP requests to arbitrary URLs by supplying user-controlled input to the skill import service (importFromUrl) and topic cover update (fetchImageFromUrl) endpoints, which use the global fetch without the project's ssrf-safe-fetch wrapper. Attackers can target internal addresses such as cloud instance metadata endpoints through these unprotected code paths to disclose internal service responses and cloud credentials. | high | 2026-07-14 |
| CVE-2026-59094 | Pathway through 0.31.1, fixed in commit d09722e, document store applies a caller-supplied glob pattern to indexed document paths using a hand-written recursive matcher that branches two ways on each ** token without memoization, giving exponential worst-case complexity. The filepath_globpattern value is taken from the body of the unauthenticated HTTP endpoints /v1/retrieve, /v1/inputs and /v2/answer and compiled into a filter evaluated once per indexed document, with no length or **-count limit. A remote unauthenticated attacker can submit a short pattern containing many ** tokens to consume CPU for tens of seconds per request, and a small number of requests denies service. | high | 2026-07-14 |
| CVE-2026-59093 | Weaviate before 1.38.0 does not verify that a principal performing an RBAC role assignment holds the permissions granted by the assigned role. The assignRoleToUser and assignRoleToGroup handlers (POST /authz/users/{id}/assign and /authz/groups/{id}/assign) authorize only that the caller may assign roles to the target user or group, not the permissions contained in the assigned roles, unlike role creation which enforces that a user can only create roles with permissions less than or equal to its own. A user holding only the delegated assign_and_revoke_users or assign_and_revoke_groups permission can assign the built-in admin role, or any high-privilege custom role, to itself or others, escalating to full administrative control of the database. | high | 2026-07-14 |
| CVE-2026-59092 | JuiceFS through 1.3.1, fixed in commit a46979c, contains an authentication bypass vulnerability that allows unauthenticated remote attackers to access sensitive debug and metrics endpoints by exploiting improper handler registration on the shared http.DefaultServeMux. Attackers can request the /debug/pprof/cmdline endpoint to obtain the process command line containing metadata engine connection strings with database credentials, granting full read/write access to filesystem metadata, while other pprof handlers leak internal state and profiling handlers enable denial of service. | high | 2026-07-14 |
| CVE-2026-58580 | LobeChat through 2.2.9 server-database deployments are vulnerable to broken object-level authorization in MessageModel. The updateMessagePlugin, updatePluginState, updatePluginError, updateTTS and updateTranslate methods filter target rows by message id alone, omitting the userId scope that sibling methods apply, and findMessagePlugin reads back by id alone. Reachable via the corresponding tRPC message procedures, an authenticated user who knows another user's message identifier can overwrite that victim's plugin tool-call metadata, plugin state/error, text-to-speech and translation records on the same instance, and the tampered content is served back to the victim. Exploitation requires knowledge of the victim's non-enumerable message identifier. | medium | 2026-07-14 |
| CVE-2026-58579 | RAGFlow before 0.26.3 stores an agent pipeline (DSL) node name without sanitization: the agent update endpoint normalizes the submitted DSL via normalize_dsl, which only performs JSON serialization validation and preserves the node name verbatim. The dataflow-result web UI then renders that name into the "Rerun from current step" confirmation modal via dangerouslySetInnerHTML, and the i18next configuration sets escapeValue:false, so the value is inserted into the DOM without HTML encoding. An authenticated workspace user who can create or edit an agent can inject arbitrary JavaScript that executes in the session of another workspace member who opens the dataflow result and clicks rerun, enabling session/token theft and account takeover across the user trust boundary. | medium | 2026-07-14 |
| CVE-2026-58578 | LobeChat before version 2.2.10-canary.15 contains a regular expression denial of service (ReDoS) vulnerability that allows authenticated attackers to block the Node.js event loop by supplying a catastrophic-backtracking pattern in a GitHub repository URL path during skill import. Attackers can craft a malicious basePath value containing unescaped regex metacharacters such as catastrophic-backtracking patterns, which are injected into a dynamically constructed regular expression in the findSkillMd function and executed synchronously against archive entries, denying service to all concurrent users for tens of seconds per request. | high | 2026-07-14 |
| CVE-2026-58467 | Cockpit CMS through 2.14.0 contains a path traversal and local file inclusion vulnerability that allows unauthenticated attackers to read arbitrary files or execute PHP files by including unvalidated PATH_INFO derived from REQUEST_URI in filesystem path construction without containment checks. Attackers can inject dot-dot sequences into the URL to traverse outside the designated spaces directory, and when the resolved path ends with a .php extension, the application passes it to include(), enabling local file inclusion on deployments using the PHP built-in server or certain non-default Nginx configurations. | high | 2026-07-14 |
| CVE-2026-58466 | AutoBangumi before 3.2.8 contains a hard-coded default credentials vulnerability that allows unauthenticated attackers to authenticate as the administrator by using the publicly known default credentials seeded at startup via add_default_user() in the database user module when the users table is empty. Attackers can submit the default credentials to the authentication login endpoint to gain full control of the application, including RSS feed configuration, downloader configuration, and all authenticated API endpoints. | critical | 2026-07-14 |
| CVE-2026-58381 | A flaw was found in GIMP's PSP file format parser. A double-free condition occurs in the read_layer_block() function when processing a specially crafted PSP file. This could allow an attacker to cause memory corruption, potentially leading to denial of service or arbitrary code execution. | medium | 2026-07-07 |
| CVE-2026-52187 | Buffer Overflow vulnerability in UTT nv518G nv518GV3v3.2.7-210919-161313 allows a remote attacker to cause a denial of service via the gohead/sub_483ba0 component | high | 2026-07-06 |
| CVE-2025-71385 | Netdata before 2.3.1 reflects the user-supplied love query parameter of the api/v2/ilove.svg and api/v3/ilove.svg endpoints verbatim into the generated SVG document (into a text element) without HTML or XML escaping, and serves the response with Content-Type image/svg+xml. An attacker can craft a URL such as /api/v2/ilove.svg?love=<script>...</script>; when a victim navigates to it the injected script executes in the victim browser in the origin of the Netdata instance (reflected cross-site scripting). These endpoints are registered with HTTP_ACL_NOCHECK and anonymous access and, because bearer-token protection is disabled by default, are reachable without authentication on a default Netdata agent. The issue was resolved by removing the ilove endpoint. | medium | 2026-07-14 |
| CVE-2026-52734 | zebrad has unbounded memory leak in mempool download pipeline via timeout path cancel_handles retention | medium | |
| CVE-2026-52733 | zebrad has persistent on-disk corruption of Sapling/Orchard subtree roots after chain fork via pop_tip | medium | |
| CVE-2026-52739 | Zebra: Repeated Non-Finalized Shielded Transaction Aborts Zebra Before Duplicate-Nullifier Rejection | medium | |
| CVE-2026-52738 | Zebra: Finalized address balance credit-first overflow on consensus-valid blocks | medium | |
| CVE-2026-52737 | Zebra has sync restart poisoning from single unauthenticated peer via above-lookahead block | medium | |
| CVE-2026-52735 | zebrad has consensus divergence via P2SH sigop undercount in pure-Rust disabled-opcode parser | critical | |
| CVE-2026-52736 | Zebra has block suppression via NU5 same-header body poisoning of sent-hash cache | high | |
| CVE-2026-52732 | zebrad has mempool transaction admission denial via single-peer inbound queue saturation | medium | |
| CVE-2026-52731 | zebrad has full node denial of service via non-ASCII LongPollId in getblocktemplate | medium | |
| CVE-2026-49255 | electerm has Command Injection in File System Operations (rmrf, mv, cp) | high | |
| CVE-2026-49254 | Dragonfly Manager OAuth provider client_secret disclosure via unauthenticated GET /api/v1/oauth | medium | |
| CVE-2026-49253 | electerm has Path Traversal in Zmodem and Trzsz Download Filename Handling | high | |
| CVE-2026-49250 | @conform-to/dom parseSubmission vulnerable to CPU exhaustion when parsing many unique form fields | high | |
| CVE-2026-7311 | The TinyPNG – JPEG, PNG & WebP image compression plugin for WordPress is vulnerable to arbitrary file deletion due to insufficient file path validation in the delete_converted_image_size function in all versions up to, and including, 3.6.13. This makes it possible for authenticated attackers, with author-level access and above, to delete arbitrary files on the server, which can easily lead to remote code execution when the right file is deleted (such as wp-config.php). An attacker can exploit this by injecting an arbitrary server file path into the 'convert.path' field of the 'tiny_compress_images' post meta on an attachment they own, then triggering attachment deletion to invoke the vulnerable code path. | high | 2026-07-06 |
| CVE-2026-58465 | Eclipse Wakaama before snapshot/2026-05-26 contains an unbounded memory allocation vulnerability in the CoAP Block1 handler within coap/block.c that allows unauthenticated remote attackers to exhaust server memory by sending a sequence of Block1 PUT requests with incrementing block numbers. Attackers can target the registration endpoint over UDP without authentication, causing the server to repeatedly reallocate a growing accumulation buffer by appending each block payload without enforcing any maximum total size limit, resulting in denial of service through memory exhaustion. | high | 2026-07-14 |
| CVE-2026-13743 | CubeSpace CW0057 Reaction Wheel firmware versions prior to 5.0.20 are vulnerable to an Improper Verification of Cryptographic Signature vulnerability. This could allow an attacker with physical access to the product to upload arbitrary malicious firmware to the device without authentication. | medium | 2026-07-06 |
| CVE-2026-49245 | SFTPGo has stored XSS via inline parameter on public shares and user file download | low | |
| CVE-2026-49244 | SFTPGo has path confinement bypass in public browsable share partial ZIP download | medium | |
| CVE-2026-50290 | @asymmetric-effort/specifyjs: CSS expression sanitization is bypassable in renderToString | medium | |
| CVE-2026-50288 | @asymmetric-effort/specifyjs: URL parse failure silently allows request | high | |
| CVE-2026-52854 | mediawiki/maps has stored XSS through the overlays parameter in the display_map parser function | high | |
| CVE-2026-50192 | Kerberos Hub private key (X-Kerberos-Hub-PrivateKey) leaked to cross-host redirect target due to redirect-following HTTP client without CheckRedirect | medium | |
