Vulnerabilities exploitable today
413,612in current view
Single score combining CVSS, KEV membership and EPSS. Every CVE with its own record — timeline from publication to active exploitation.
In KEV catalog1,779
New KEV · 24H0
Exploit Today ≥ 701,707
Distribution · last window
- Critical5,728
- High22,707
- Medium21,208
- Low2,163
Filters
Window
Severity
Flags
CVECVSSEPSSKEVRExploitTitleMod.
CVE-2026-1262528.3 HIG—
———A improper neutralization of special elements used in an sql command ('sql injection') vulnerability in Fortinet FortiAuthenticator 8.0.0 through 8.0.3, FortiAuthenticator 6.6.0 through 6.6.10 may allow an authenticated attacker to execute unauthorized code or commands via http requests11hCVE-2026-1262377.5 HIG—
———A loop with unreachable exit condition ('infinite loop') vulnerability in Fortinet FortiAuthenticator 8.0.0 through 8.0.3, FortiAuthenticator 6.6.0 through 6.6.10 may allow an attacker to cause a denial of service via crafted TCP requests.11hCVE-2026-1262299.1 CRI—
———A authorization bypass through user-controlled key vulnerability in Fortinet FortiAnalyzer 8.0.0 through 8.0.1, FortiAnalyzer 7.6.0 through 7.6.7, FortiAnalyzer 7.4.0 through 7.4.11, FortiAnalyzer 7.2 all versions, FortiAnalyzer Cloud 8.0.1, FortiAnalyzer Cloud 7.6.0 through 7.6.7, FortiAnalyzer Cloud 7.4.1 through 7.4.11, FortiAnalyzer Cloud 7.2 all versions, FortiManager 8.0.0 through 8.0.1, FortiManager 7.6.0 through 7.6.7, FortiManager 7.4.0 through 7.4.11, FortiManager 7.2 all versions, FortiManager Cloud 8.0.1, FortiManager Cloud 7.6.0 through 7.6.7, FortiManager Cloud 7.4.1 through 7.4.11, FortiManager Cloud 7.2 all versions may allow an authenticated administrator to access to other VDOMs via specially crafted HTTP requests.11hCVE-2026-1262267.2 HIG—
———A improper neutralization of special elements used in an os command ('os command injection') vulnerability in Fortinet FortiAnalyzer 8.0.0, FortiAnalyzer 7.6.0 through 7.6.7, FortiAnalyzer 7.4.0 through 7.4.11, FortiAnalyzer 7.2.0 through 7.2.12, FortiAnalyzer Cloud 8.0.0, FortiAnalyzer Cloud 7.6.0 through 7.6.7, FortiAnalyzer Cloud 7.4.1 through 7.4.11, FortiAnalyzer Cloud 7.2.1 through 7.2.12, FortiManager 8.0.0, FortiManager 7.6.0 through 7.6.7, FortiManager 7.4.0 through 7.4.11, FortiManager 7.2.0 through 7.2.12, FortiManager Cloud 8.0.0, FortiManager Cloud 7.6.0 through 7.6.7, FortiManager Cloud 7.4.1 through 7.4.11, FortiManager Cloud 7.2.1 through 7.2.12 may allow an authenticated privileged attacker to execute unauthorized code or commands via crafted report schedule name.11hCVE-2026-1262237.2 HIG—
———A stack-based buffer overflow vulnerability in Fortinet FortiAnalyzer 8.0.0 through 8.0.1, FortiAnalyzer 7.6.0 through 7.6.7, FortiAnalyzer 7.4.0 through 7.4.11, FortiAnalyzer Cloud 8.0.0 through 8.0.1, FortiAnalyzer Cloud 7.6.0 through 7.6.7, FortiAnalyzer Cloud 7.4.1 through 7.4.11, FortiManager 8.0.0 through 8.0.1, FortiManager 7.6.0 through 7.6.7, FortiManager 7.4.0 through 7.4.11, FortiManager Cloud 8.0.0 through 8.0.1, FortiManager Cloud 7.6.0 through 7.6.7, FortiManager Cloud 7.4.1 through 7.4.11 may allow an authenticated administrator to execute unauthorized code or commands or cause a denial of service via specially crafted JSON-RPC requests.11hCVE-2026-1262217.2 HIG—
———A improper neutralization of special elements in output used by a downstream component ('injection') vulnerability in Fortinet FortiAnalyzer 8.0.0, FortiAnalyzer 7.6.0 through 7.6.7, FortiAnalyzer 7.4.0 through 7.4.11, FortiAnalyzer Cloud 8.0.0, FortiAnalyzer Cloud 7.6.0 through 7.6.7, FortiAnalyzer Cloud 7.4.1 through 7.4.11, FortiManager 8.0.0, FortiManager 7.6.0 through 7.6.7, FortiManager 7.4.0 through 7.4.11, FortiManager Cloud 8.0.0, FortiManager Cloud 7.6.0 through 7.6.7, FortiManager Cloud 7.4.1 through 7.4.11 may allow an authenticated administrator to read or write arbitrary files or execute unauthorized code or commands via specially crafted requests to the GUI.11hCVE-2026-1262158.2 HIG—
———A incorrect authorization vulnerability in Fortinet FortiAnalyzer 8.0.0 through 8.0.1, FortiAnalyzer 7.6.0 through 7.6.7, FortiAnalyzer 7.4.0 through 7.4.11, FortiAnalyzer 7.2.0 through 7.2.12, FortiAnalyzer Cloud 8.0.0 through 8.0.1, FortiAnalyzer Cloud 7.6.0 through 7.6.7, FortiAnalyzer Cloud 7.4.1 through 7.4.11, FortiAnalyzer Cloud 7.2.1 through 7.2.12, FortiManager 8.0.0 through 8.0.1, FortiManager 7.6.0 through 7.6.7, FortiManager 7.4.0 through 7.4.11, FortiManager 7.2.0 through 7.2.12, FortiManager Cloud 8.0.0 through 8.0.1, FortiManager Cloud 7.6.0 through 7.6.7, FortiManager Cloud 7.4.1 through 7.4.11, FortiManager Cloud 7.2.1 through 7.2.12 may allow an authenticated administrator to escalate their privileges, bypass authorization controls or impersonate internal services via specially crafted requests to the GUI.11hCVE-2026-1262137.1 HIG—
———A improper neutralization of special elements used in an sql command ('sql injection') vulnerability in Fortinet FortiAnalyzer 8.0.0 through 8.0.1, FortiAnalyzer 7.6.0 through 7.6.7, FortiAnalyzer 7.4 all versions, FortiAnalyzer 7.2 all versions, FortiAnalyzer Cloud 8.0.0 through 8.0.1, FortiAnalyzer Cloud 7.6.0 through 7.6.7, FortiAnalyzer Cloud 7.4.1 through 7.4.11, FortiAnalyzer Cloud 7.2.1 through 7.2.12, FortiManager 8.0.0 through 8.0.1, FortiManager 7.6.0 through 7.6.7, FortiManager 7.4.0 through 7.4.11, FortiManager 7.2.0 through 7.2.12, FortiManager Cloud 8.0.0 through 8.0.1, FortiManager Cloud 7.6.0 through 7.6.7, FortiManager Cloud 7.4.1 through 7.4.11, FortiManager Cloud 7.2.1 through 7.2.12 may allow an authenticated attacker to execute unauthorized SQL commands via specially crafted requests.11hCVE-2026-1261797.2 HIG—
———A improper neutralization of special elements used in an os command ('os command injection') vulnerability in Fortinet FortiAnalyzer 8.0.0, FortiAnalyzer 7.6.0 through 7.6.7, FortiAnalyzer 7.4.0 through 7.4.11, FortiAnalyzer 7.2.0 through 7.2.12, FortiAnalyzer Cloud 8.0.0, FortiAnalyzer Cloud 7.6.0 through 7.6.7, FortiAnalyzer Cloud 7.4.1 through 7.4.11, FortiAnalyzer Cloud 7.2.1 through 7.2.12, FortiManager 8.0.0, FortiManager 7.6.0 through 7.6.7, FortiManager 7.4.0 through 7.4.11, FortiManager 7.2.0 through 7.2.12, FortiManager Cloud 8.0.0, FortiManager Cloud 7.6.0 through 7.6.7, FortiManager Cloud 7.4.1 through 7.4.11, FortiManager Cloud 7.2.1 through 7.2.12 may allow an authenticated administrator to execute unauthorized code or commands via specially crafted requests11hCVE-2026-1261687.2 HIG—
———A improper neutralization of argument delimiters in a command ('argument injection') vulnerability in Fortinet FortiMail 8.0.0 through 8.0.2, FortiMail 7.6.0 through 7.6.6, FortiMail 7.4.0 through 7.4.8, FortiMail 7.2 all versions may allow an authenticated remote attacker with domain admin or admin privileges to delete users of other tenant domains on a paired FortiAuthenticator, execute arbitrary code as root or arbitrary CLI commands, and read or write arbitrary files as root via crafted HTTP requests to the administrative interface.11hCVE-2026-1261559.1 CRI—
———A incorrect authorization vulnerability in Fortinet FortiMail 8.0.0 through 8.0.1, FortiMail 7.6.0 through 7.6.6, FortiMail 7.4.0 through 7.4.8, FortiMail 7.2 all versions may allow an unauthenticated remote attacker to obtain an authenticated IBE webmail session as any IBE recipient and read or send their encrypted mail via crafted HTTP requests.11hCVE-2026-1215288.1 HIG—
———A improper certificate validation vulnerability in Fortinet FortiOS 7.6.0 through 7.6.4, FortiOS 7.4.0 through 7.4.9, FortiOS 7.2.0 through 7.2.13 may allow a remote and unauthenticated attacker to perform a Man-in-the-Middle attack on the communication channel between the SSLVPN web portal and the backend destination website.11hCVE-2026-65965——
———ideon is a self-hosted visual workspace. Prior to version 0.9.4, several project mutation endpoints under /api/projects/[id]/* resolve the caller's project role but never enforce it, so a user who was granted only the read-only viewer role on a project can perform write operations on that project. The most severe is POST /api/projects/[id]/graph, which deletes every block and link in the project and re-inserts client-supplied ones. A viewer can silently rewrite or wipe the entire canvas of a project they were invited to read. The application clearly intends viewer to be read-only: the collaborators endpoint, the automation endpoints, and the realtime Yjs sync path all reject write attempts from viewers. These specific HTTP handlers simply omit the guard their siblings enforce. This issue has been patched in version 0.9.4.12hCVE-2026-65957——
———Helio is an MCP proxy that sits between AI agents and the tools they use. Prior to version 0.8.0, the dashboard sideband (default 127.0.0.1:3100) validated CORS request origins by matching the hostname against a private-network string prefix (10., 172.16.–172.31., 192.168.) rather than validating an IP address. A public DNS name such as 192.168.attacker.com, 10.evil.io, or 172.16.pwn.example also satisfies that prefix, so the proxy echoed the attacker origin in Access-Control-Allow-Origin. This issue has been patched in version 0.8.0.12hCVE-2026-1268136.8 MED—
———Roundcube Webmail before 1.6.21 and 1.7.x before 1.7.6 contains an IMAP command injection vulnerability that allows attackers to inject commands via CRLF in the _uid parameter when messages_cache is enabled. Attackers can lure logged-in victims into opening crafted GET links to mail/show or mail/get, executing IMAP commands that delete messages, change flags, or grant mailbox access via SETACL.12hCVE-2026-1268115.4 MED—
———BookStack before 26.09.3 contains an incorrect authorization vulnerability that allows authenticated users to access other users' unpublished draft page images by checking only PageView permission without draft status. Attackers can read drawio data via the base64 endpoint or rename, replace, or delete draft images by sequential ID to disclose or tamper with unpublished work.12hCVE-2026-1268097.1 HIG—
———WWBN AVideo through 29.7.1 contains a missing authorization vulnerability in objects/import.json.php that allows uploaders to import other users' MP4 files by referencing any folder under videos/. Attackers can copy private, unlisted, password-protected or pay-per-view videos, read victims' .txt/.html sidecars as descriptions, list files via objects/listFiles.json.php, and delete victims' media using the delete parameter.12hCVE-2026-1268068.7 HIG—
———WWBN AVideo 14.3 through 29.7.1 contains a stored cross-site scripting vulnerability that allows uploaders to inject scripts via the videoDownloadedLink value stored unescaped through objects/aVideoEncoder.json.php. Attackers can embed event-handler payloads in the URL host rendered as a Source tag in view/managerVideos_body.php, executing in administrator sessions to install plugin archives containing PHP.12hCVE-2026-1268048.1 HIG—
———WWBN AVideo 12.4 through 29.7.1 contains a path traversal vulnerability in the ADs plugin deleteImage.json.php that allows authenticated users with custom ads to delete arbitrary files. Attackers can set the unvalidated is_regular_user parameter to another user's id or ../ sequences to delete other users' ads or files like videos/configuration.php, taking the site offline and re-exposing the installer.12hCVE-2026-1267997.5 HIG—
———Gotenberg 8.31.0 before 8.38.0 contains a server-side request forgery vulnerability in IsPublicIP() that allows unauthenticated attackers to bypass private-IP deny-list protections using IPv6 zone identifiers. Attackers can append a zone such as %25eth0 to the unspecified address or blocked IPv6 prefixes in downloadFrom, webhook, or Chromium/LibreOffice URLs to reach internal services and read responses in the returned PDF.12hCVE-2026-103980——
———ideon is a self-hosted visual workspace. Prior to version 0.9.6, the SSRF protection in safeFetch is bypassed by IPv6 transition addresses that embed private IPv4 addresses. The isPrivateIP function in src/app/lib/ssrf.ts checks only a narrow set of IPv6 ranges (loopback ::1, ULA fc/fd, link-local fe80) but does not recognize IPv4-mapped (::ffff:), NAT64 well-known (64:ff9b::/96, RFC 6052), NAT64 local-use (64:ff9b:1::/48, RFC 8215), 6to4 (2002::/16, RFC 3056), or Teredo (2001::/32, RFC 4380) encodings. An authenticated user can supply these addresses in link URLs to reach cloud metadata endpoints (169.254.169.254), localhost services, and RFC 1918 internal networks through the link preview metadata fetcher. This issue has been patched in version 0.9.6.12hCVE-2026-911517.3 HIG—
———A flaw was found in Cockpit. A local authenticated user with control over a target account's home directory can exploit a privileged symlink dereference vulnerability in the `ssh-add-public-key.sh` helper. When this helper runs with elevated privileges, it can be tricked into following attacker-controlled symlinks within the `.ssh` or `authorized_keys` paths. This allows the attacker to arbitrarily change file ownership and modes, potentially leading to local privilege escalation and impacting the integrity, confidentiality, and availability of the system.13hCVE-2026-0328——
———Authentication bypass vulnerabilities in the GlobalProtect™ portal and gateway of Palo Alto Networks PAN-OS® software enables an unauthenticated user to bypass security restrictions and establish an unauthorized VPN connection.
Panorama, Cloud NGFW, and Prisma® Access are not affected by this vulnerability.13hCVE-2026-918808.9 HIG—
———Dell Terraform Provider for PowerFlex, versions 1.5.0 through 1.8.0, contains an Improper Certificate Validation vulnerability in the SSH client component.Terraform PowerFlex provider uses ssh.InsecureIgnoreHostKey() as the default SSH host key callback. An unauthenticated attacker with remote access could potentially exploit this vulnerability, leading to information exposure and unauthorized access.14hCVE-2026-918738.7 HIG—
———Dell OpenManage Ansible Modules, versions 2.1.2 through 10.0.3, contains an Improper Certificate Validation vulnerability in the iDRACRedfishAPI component. An unauthenticated attacker with remote access could potentially exploit this vulnerability, leading to information exposure and unauthorized access.14hCVE-2026-760783.1 LOW—
———Probo is a self-hostable governance, risk, and compliance (GRC) platform built for engineering and security teams. Prior to version 0.224.0, the public Trust Center GraphQL API exposes acceptElectronicSignature and recordSigningEvent, both gated only by @authentication(required: PRESENT) @sessionOnly — any self-provisioned trust center identity can call them. The underlying esign.Service.AcceptSignature and esign.Service.recordEvent build the tenant scope from the client-supplied signatureId GID (coredata.NewScopeFromObjectID(req.SignatureID)) and load the signature by that GID. At no point do they verify that the signature belongs to the caller or that the caller's email matches the signature's SignerEmail. As a result, any authenticated trust-center visitor who possesses another visitor's signature GID can: (1) complete that visitor's NDA signature — overwriting signer name/IP/UA with attacker data and setting status to Accepted, which triggers the sealing worker to generate a completion certificate; (2) inject arbitrary audit-trail events into any signature's log. Both attacks defeat the non-repudiation and integrity guarantees of the e-signature system. This issue has been patched in version 0.224.0.14hCVE-2026-659677.1 HIG—
———Fider is a feedback portal for feature requests and suggestions. Prior to version 0.34.0, a Server-Side Request Forgery (SSRF) vulnerability exists in Fider due to insufficient validation of webhook URLs. The application allows user-controlled URLs to be fetched by the server without restricting access to internal or external network resources. This enables attackers with administrative privileges to make arbitrary HTTP requests from the server, potentially accessing internal services or cloud metadata endpoints. This issue has been patched in version 0.34.0.14hCVE-2026-65833——
———iopress is a native HTTP server for Node.js with platform-specific async I/O. Prior to version 1.0.5, the chunked-transfer-encoding HTTP request parser in @iopress/core reads past the end of the input buffer when a request declares a chunk size larger than the body actually received. This results in a heap-buffer-over-read. This issue has been patched in version 1.0.5.14hCVE-2026-585057.5 HIG—
———geojson-vt is a JavaScript library for slicing GeoJSON data into vector tiles on the fly, primarily designed to enable rendering and interacting with large geospatial datasets on the browser side (without a server). Prior to version 6.1.1, @maplibre/geojson-vt contains an uncontrolled recursion vulnerability in its GeoJSON conversion logic. When a GeoJSONVT instance is constructed with a deeply nested GeometryCollection object, the mutually recursive functions featureToInternal and convertGeometryCollection in src/convert.ts exhaust the Node.js call stack, causing an unhandled RangeError: Maximum call stack size exceeded that terminates the process. Because the library is widely used in server-side tile services that accept GeoJSON over the network, a remote unauthenticated attacker can trigger a denial-of-service by sending a single crafted payload. No authentication is required, and no non-default library options are needed. This issue has been patched in version 6.1.1.14hCVE-2026-1267807.2 HIG—
———Webmin through 2.670 contains a privilege escalation vulnerability in acl/save_group.cgi that allows delegated administrators to grant unauthorized modules by submitting arbitrary module lists in the mod parameter. Attackers with Webmin Users module access can create or edit groups holding modules like Command Shell and assign themselves membership to execute commands as root.14hCVE-2026-1267724.3 MED—
———OpsTree REMS through 1.0.0 contains an improper attribution vulnerability that allows authenticated users to post comments as the incident creator via POST /api/v1/incidents/{incident_id}/comment. Lower-privileged attackers can comment on incidents owned by users such as a Super User to impersonate them and corrupt the incident timeline and audit trail.14hCVE-2026-1238776.8 MED—
———A certificate validation bypass in Node.js https.Agent lets a cached TLS session authenticated under one CA policy be reused by a later request with a stricter policy. https.Agent.prototype.getName() assigns the same Agent and session-cache key to a request that omits ca and one that explicitly supplies ca: [], because an omitted value appends nothing to the key while an empty array stringifies to an empty string. The two policies differ: omitting ca uses the default trust store, while supplying ca replaces it. Since they share a key, a TLS session authenticated under the broader default policy is handed to a new connection made under the later ca: [] policy, and TLS 1.3 session resumption carries the earlier authentication forward, so the second request reports authorized === true and transmits confidential data to a server its own CA policy would reject on a fresh connection. This vulnerability affects Node.js 26.x, 24.x, and 22.x.14hCVE-2026-1238674.2 MED—
———A per-host mutual-TLS authorization bypass in Node.js lets a client authorized for the default hostname reach a protected hostname it is not authorized for. When a TLS server assigns a hostname a different client-certificate CA with server.addContext(host, { ca: protectedCA }) but no server certificate in that SNI context, the per-host client CA is never installed and the handshake falls back to the default client CA. In src/crypto/crypto_tls.cc, CertCbDone installs the SNI context's CA only when setSniContext succeeds, but setSniContext returns false for a context with no server certificate, so a CA-only SNI context silently keeps the default CA. A client holding a certificate trusted by the default CA, but not the protected CA, can therefore complete mTLS to the protected hostname and read its response even with requestCert and rejectUnauthorized enabled, while a client certified by the protected CA is rejected. This vulnerability affects Node.js 26.x, 24.x, and 22.x.14hCVE-2026-1238667.5 HIG—
———A use-after-free in Node.js HTTP/2 stream teardown can leak response memory, including the contents of other connections' response buffers, to a peer. When a stream is torn down while a write is still backpressured, for example by a peer RST_STREAM or by the server's own per-stream timeout destroying the stream, the stream's response buffer is freed but a reference to it is still used, so freed or unrelated heap memory is written to the socket instead of the intended response. On Node.js 26.x and 24.x the teardown can be driven purely by a remote peer once the window-size change from 26.9.0 and 24.22.0 is present, and on all supported lines the same defect is reachable through an application-initiated teardown such as a per-stream idle timeout. This vulnerability affects Node.js 26.x, 24.x, and 22.x.14hCVE-2026-1238638.2 HIG—
———A reachable assertion in Node.js HTTP/2 session teardown lets any unauthenticated client abort the server, over both TLS and cleartext. A client can arrange for an Http2Stream that still holds nghttp2-allocated header buffers to outlive its session: the stream is removed from the session's stream set but never detached, so when ~Http2Session() runs the memory accounting is nonzero and the CHECK_EQ(current_nghttp2_memory_, 0) assertion fails, terminating the process. The trigger needs only a default http2 server whose handler responds before the request body has ended, which is typical. It is a variant of CVE-2024-27983 that the earlier fix did not cover, because that fix only detached streams still tracked in the session. This vulnerability affects Node.js 26.x (since 26.9.0) and 24.x (since 24.22.0).14hCVE-2026-1238292.5 LOW—
———A Permission Model bypass in Node.js recursive file watching discloses the names of files and directories the process is not permitted to read. On macOS and Windows, fs.watch() and fs.promises.watch() in recursive mode are backed by the native libuv watcher: FSEventWrap::Start checks kFileSystemRead only once, on the watched root, and FSEventWrap::OnEvent then forwards every filename libuv reports (relative paths at any depth) to the JavaScript callback without checking that each path is covered by the read grant. Code running under --permission with an exact read grant on a directory can therefore learn names such as sub/secret.txt even though readdir on that subdirectory is denied. The flaw exposes names and timing metadata, not file contents, and requires macOS or Windows plus files created under the watched directory by another actor. This vulnerability affects Node.js 26.x, 24.x, and 22.x.14hCVE-2026-1238285.9 MED—
———A one-byte heap buffer overflow in Node.js HTTP/2 Alt-Svc handling lets a remote, unauthenticated attacker corrupt process memory. When an application passes an attacker-controlled value such as a request's :authority header to Http2Session.altsvc(), the AltSvc implementation in src/node_http2.cc allocates a buffer sized to the origin length and then writes it with a null terminator, placing one byte past the end of the allocation. A single request carrying a long but syntactically valid :authority triggers the out-of-bounds write, which aborts the process and serves as a remote heap-corruption primitive. This vulnerability affects Node.js 26.x and 24.x.14hCVE-2026-648585.3 MED—
———vulnerability-operator scans SBOMs and Images for vulnerabilities. Prior to version 0.28.16, an unauthenticated information-disclosure vulnerability exposes all vulnerability scan reports served by the operator. The /report/ HTTP endpoint mounts a raw http.FileServer over the reports directory (default /reports) with no authentication, authorization, or middleware. Any client that can reach the operator's HTTP port (default 8080) can list and download every JSON vulnerability report, disclosing CVE IDs, affected package names/versions, and container image identifiers for every workload scanned in the cluster. This issue has been patched in version 0.28.16.15hCVE-2026-1085606.1 MED—
———A reflected cross-site scripting vulnerability in updates-history.php in Revive Adserver 6.0.8 and earlier allows an unauthenticated attacker to execute JavaScript in a victim's browser by persuading them to visit a crafted URL. Attacker-controlled values are included in XML command attributes without appropriate encoding, allowing injection of executable XHTML into the response. The preliminary session check omits GET xajax requests, exposing an unauthenticated attack path, while authenticated administrator requests also reach the unsafe output. The issue is fixed in Revive Adserver 6.0.9 by applying the session check to both GET and POST requests and encoding XML command attribute values in both response serializers.15hCVE-2026-122043——
———repmgr 4.0.0 through 5.5.0 is affected by an OS command injection vulnerability (CWE-78) in event notifications. When event_notification_command is configured, repmgr substitutes the %a (node name), %c (connection string) and %d (details) placeholders into the command and runs it through the shell. %a and %c are inserted without escaping, and %d escapes only double quotes, so shell command substitution still works inside the quoting the documentation recommends. The node name and connection string come from node records in the repmgr.nodes table. An attacker with the credentials of a non-superuser repmgr database user can write a crafted node record and run OS commands, as the operating system user that runs repmgr or repmgrd (usually postgres), on other cluster nodes that process an event using these placeholders. The %a placeholder of failover_validation_command has the same flaw. Deployments whose repmgr database user is a superuser gain no new capability, because a superuser can already run OS commands, and deployments that do not use these placeholders are not affected. The issue is fixed in repmgr 5.5.1. After upgrading, operators must remove any quotes they placed around these placeholders, because a placeholder left inside double quotes is still open to command injection.16h