The frontend gRPC server's streaming interceptor chain did not include the authorization interceptor. When a ClaimMapper and Authorizer are configured, unary RPCs enforce authentication and authorization, but the streaming AdminService/StreamWorkflowReplicationMessages endpoint accepted requests without credentials. This endpoint is registered on the same port as WorkflowService and cannot be disabled independently. An attacker with network access to the frontend port could open the replication stream without authentication. Data exfiltration is possible, but  only when a configured replication target is correctly configured and the attacker has knowledge of the cluster configuration, as the history service validates cluster IDs and peer membership before returning replication data. Temporal Cloud is not affected.

OpenClaw before 2026.3.24 contains a privilege escalation vulnerability where the /allowlist command fails to re-validate gateway client scopes for internal callers, allowing operator.write-scoped clients to mutate channel authorization policy. Attackers can exploit chat.send to build an internal command-authorized context and persist channel allowFrom and groupAllowFrom policy changes reserved for operator.admin scope.

OpenClaw before 2026.3.24 contains an incomplete fix for CVE-2026-27486 where the !stop chat command uses an unpatched killProcessTree function from shell-utils.ts that sends SIGKILL immediately without graceful SIGTERM shutdown. Attackers can trigger process termination via the !stop command, causing data corruption, resource leaks, and skipped security-sensitive cleanup operations.

Rembg is a tool to remove images background. Prior to 2.0.75, a path traversal vulnerability in the rembg HTTP server allows unauthenticated remote attackers to read arbitrary files from the server's filesystem. By sending a crafted request with a malicious model_path parameter, an attacker can force the server to attempt loading any file as an ONNX model, revealing file existence, permissions, and potentially file contents through error messages. This vulnerability is fixed in 2.0.75.

LiteLLM through 2026-04-08 allows remote attackers to execute arbitrary code via bytecode rewriting at the /guardrails/test_custom_code URI.

Vikunja is an open-source self-hosted task management platform. Prior to 2.3.0, the addRepeatIntervalToTime function uses an O(n) loop that advances a date by the task's RepeatAfter duration until it exceeds the current time. By creating a repeating task with a 1-second interval and a due date far in the past, an attacker triggers billions of loop iterations, consuming CPU and holding a database connection for minutes per request. This vulnerability is fixed in 2.3.0.

OpenClaw before 2026.3.24 contains a path traversal vulnerability in sandbox enforcement allowing sandboxed agents to read arbitrary files from other agents' workspaces via unnormalized mediaUrl or fileUrl parameter keys. Attackers can exploit incomplete parameter validation in normalizeSandboxMediaParams and missing mediaLocalRoots context to access sensitive files including API keys and configuration data outside designated sandbox roots.

Apache Log4cxx's XMLLayout https://logging.apache.org/log4cxx/1.7.0/classlog4cxx_1_1xml_1_1XMLLayout.html , in versions before 1.7.0, fails to sanitize characters forbidden by the XML 1.0 specification https://www.w3.org/TR/xml/#charsets in log messages, NDC, and MDC property keys and values, producing invalid XML output. Conforming XML parsers must reject such documents with a fatal error, which may cause downstream log processing systems to drop or fail to index affected records. An attacker who can influence logged data can exploit this to suppress individual log records, impairing audit trails and detection of malicious activity. Users are advised to upgrade to Apache Log4cxx 1.7.0, which fixes this issue.

OpenClaw before 2026.3.22 contains a webhook reply delivery vulnerability that allows attackers to rebind chat replies to unintended users by exploiting mutable username matching instead of stable numeric user identifiers. Attackers can manipulate username changes to redirect webhook-triggered replies to different users, bypassing the intended recipient binding recorded in webhook events.

HDF5 is software for managing data. In 1.14.1-2 and earlier, an attacker who can control an h5 file parsed by HDF5 can trigger a write-based heap buffer overflow condition in the H5T__ref_mem_setnull method. This can lead to a denial-of-service condition, and potentially further issues such as remote code execution depending on the practical exploitability of the heap overflow against modern operating systems.