CVE-2026-55885

CVE-2026-55885 (Grav) : An authenticated admin with backup permissions can download a ZIP of the Grav installation via the backup/download endpoint, exposing sensitive data: admin.yaml with the administrator password hash and site configuration under user/config. The vulnerability stems from (1) ...

CVE-2026-53653

Grav (file-based web platform) contains CVE-2026-53653: unauthenticated denial of service by requesting image derivatives with oversized dimensions via URL actions like forceResize in Grav::fallbackUrl. The root cause is that request parameters are passed to ImageMedium magic actions without a di...

CVE-2026-54919

cpp-httplib on Mbed TLS and wolfSSL backends was vulnerable to a TLS certificate chain verification bypass for IP-literal hosts (affecting v0.31.0–0.46.1) when server verification was enabled; SSLClient/Client could skip chain validation and Mbed TLS WebSocketClient could skip verification. The i...

CVE-2026-49865

CVE-2026-49865 is reserved; connected advisory GHSA-pj8j-p4g4-4vw8 details a real vulnerability in Kimai 2.56.0 where invoice PDF rendering can trigger server-side requests via Markdown image URLs, enabling SSRF when attacker-controlled Markdown is rendered into invoices. The flaw occurs during t...

CVE-2026-39903

The CVE concerns Simple Machines Forum (SMF) 2.1 prior to 2.1.8 and 3.0 prior to 3.0 Alpha 5. A single-character operator error in Sources/Actions/AttachmentApprove.php causes the permission check to always pass, enabling an authenticated low-privilege user to approve, reject, or delete any pendi...

CVE-2026-59180

Apprise before v1.11.0 vulnerable: HTTP-based notification plugins and HTTP loaders (apprise/attachment/http.py, apprise/config/http.py) follow redirects and resend configured auth headers and query parameters. This can leak secrets (Authorization headers, bearer tokens, service keys) to a compro...

CVE-2026-55687

CVE-2026-55687 describes a stack-based out-of-bounds write in the Espressif ESF-IDF JPEG decoder marker parsing (jpeg_parse_dqt_marker). The attack uses the attacker-controlled DQT marker Tq nibble as an index into the qt_tbl array without validating it is within 0..3, enabling malformed JPEG inp...

CVE-2026-54063

Excelize CVE-2026-54063: The library vulnerable due to checkSheet() using attacker-controlled row attribute to allocate memory for xlsxRow slices without enforcing the Excel row limit. This can cause (A) out-of-memory leading to process kill and (B) runtime panic from out-of-bounds indexing when ...

CVE-2026-59162

CVE-2026-59162 affects the Go library Excelize (reading/writing Excel spreadsheets). Before 2.11.0, shared-string values are parsed with strconv.Atoi and only the upper bound is checked when indexing the shared string slice, which allows a -1 index to be used (sharedStrings[-1]) and causes a pani...

CVE-2026-59161

CVE-2026-59161 affects the Go library Excelize. The streaming worksheet reader used by Rows/GetRows could bypass the TotalRows limit, enabling an attacker-controlled index to cause unbounded memory and CPU usage by generating empty rows beyond 1048576. The issue is fixed in version 2.11.0; upgrad...

CVE-2026-59154

The CVE affects Wekan (open‑source Kanban) prior to version 9.64. A cross‑board authorization bypass exists in the direct Meteor collection allow rules for Checklists and ChecklistItems: updates are authorized only against the current source doc.cardId and do not inspect destination cardId or boa...

CVE-2026-15376

CVE-2026-15376 affects Eleveo Call Recording Software 9.7.0. The vulnerability targets an unknown function in /callrec/statisticReportAction.do, leading to improper authorization that can be exploited remotely. Public exploit details exist. Metrics (VulDB) show CVSS scores around 5.3–6.5 (MEDIUM)...

CVE-2026-53657

CVE-2026-53657 affects Lima when using the qemu driver. Prior to 2.1.3, an unprivileged user inside a Lima QEMU guest could access the /run/lima-guestagent.sock guest-agent socket when enabled, enabling commands with root privileges inside the VM via the socket’s tunneling to privileged daemons. ...

CVE-2026-56675

9router prior to version 0.5.2 validates loopback requests as trusted, allowing unauthenticated access to /v1/* endpoints when a reverse proxy on localhost (127.0.0.1) forwards traffic. A remote attacker could reach /v1 APIs (e.g., /v1/models) without an API key, potentially abusing configured up...

CVE-2026-55638

9router prior to version 0.5.2 exposed an unauthenticated access path via /codex that bypassed the API-key gate due to an incomplete rewrite handling. An attacker could send requests to /codex/* and trigger upstream provider calls using operator-stored LLM provider credentials. The vulnerability ...

CVE-2026-55641

9router (affected component: /v1 LLM proxy logic) had an unauthenticated bypass due to client-controlled Host header processing before version 0.5.2. This allowed a remote attacker to craft Host: localhost to bypass API-key checks, potentially enabling upstream provider calls with stored credenti...

CVE-2026-56676

9router is affected by an image prefetch DNS rebinding vulnerability (CVE-2026-56676). Prior to v0.5.2, the application validates the image URL by DNS resolution but then performs a server-side fetch with a separate DNS lookup, enabling an authenticated attacker with access to the LLM proxy to tr...

CVE-2026-55500

CVE-2026-55500 (9router) exposes a critical flaw in the /api/settings/database endpoint. The GET export returns the entire database, including plaintext API keys, OAuth tokens, and provider credentials, while POST can wipe-and-replace the database with attacker-controlled data. This bypasses auth...

CVE-2026-15143

CVE-2026-15143 affects guardrails-detectors' file_type content detector. An attacker can supply an arbitrary XML Schema (XSD) string, processed without proper restrictions, enabling server-side requests to arbitrary URLs or local file reads and risking disclosure of sensitive information.

CVE-2026-55501

The CVE concerns 9router prior to version 0.4.80, where the login rate limiter uses client identity from the attacker-controlled X-Forwarded-For header in src/lib/auth/loginLimiter.js. This allows remote attackers to rotate X-Forwarded-For per attempt, creating a fresh rate-limit bucket each time...