CVE-2026-62208
MediumCVSS 6.5Exploitation Probability (EPSS)
Low risk17th percentile - higher than 17% of all known CVEs
Summary
A vulnerability in OpenClaw before version 2026.6.5 allows forwarding of Authorization headers during MCP SSE redirects. When the feature is enabled and reachable, a lower-trust caller or configured input path can execute or persist actions beyond the caller's intended authorization.
Risk Assessment
The risk involves potential privilege escalation by an unauthorized user, leading to unauthorized access to resources or execution of administrative operations.
Recommendation
Immediately update OpenClaw to version 2026.6.5 or later and review the configuration to ensure lower-trust input cannot reach the vulnerable path.
Other vulnerabilities in OpenClaw
See all- CVE-2026-45006High
OpenClaw before version 2026.4.23 contains an improper access control vulnerability in the gateway tool's config.apply and config.patch operations, allowing compromised models to write unsafe configuration changes by bypassing an incomplete denylist protection.
- CVE-2026-45004High
OpenClaw before 2026.4.23 contains an arbitrary code execution vulnerability in the bundled plugin setup resolver that loads setup-api.js from process.cwd() during provider setup metadata resolution. Attackers can execute arbitrary JavaScript under the current user account by placing a malicious extensions/<plugin>/setup-api.js file in a repository.
- CVE-2026-45001High
OpenClaw before version 2026.4.20 contains a guard bypass vulnerability in the config.patch and config.apply endpoints. This vulnerability fails to protect operator-trusted settings such as sandbox policy, plugin enablement, gateway auth/TLS, and MCP server configuration.
- CVE-2026-44995High
OpenClaw before version 2026.4.20 contains an improper environment variable validation vulnerability in MCP stdio server configuration that allows attackers to execute arbitrary code. Malicious workspace configurations can pass dangerous startup variables to MCP server processes.
- CVE-2026-44118High
OpenClaw before version 2026.4.22 derives loopback MCP owner context from spoofable server-issued bearer tokens in request headers. Non-owner loopback clients can present themselves as owner to bypass owner-gated operations.
- CVE-2026-44115High
OpenClaw before version 2026.4.22 contains an exec allowlist analysis vulnerability allowing shell expansion hiding in unquoted heredoc bodies. Attackers can bypass allowlist validation by embedding shell expansion tokens in heredoc bodies to execute unapproved commands at runtime.
- CVE-2026-44114High
OpenClaw before version 2026.4.20 fails to properly reserve the OPENCLAW_ runtime-control environment namespace in workspace dotenv files, allowing attackers to override critical runtime variables. Malicious workspaces can set variables like OPENCLAW_GIT_DIR to manipulate trusted OpenClaw runtime behavior during source-update or installer flows.
- CVE-2026-44110High
OpenClaw before version 2026.4.15 contains an authorization bypass vulnerability in Matrix room control-command authorization. Attackers with DM-paired sender IDs can execute room control commands without being in configured allowlists.
- CVE-2026-43585High
OpenClaw before version 2026.4.15 captures resolved bearer-auth configuration at startup, allowing revoked tokens to remain valid after SecretRef rotation. Gateway HTTP and WebSocket handlers fail to re-resolve authentication per-request.
- CVE-2026-43584High
OpenClaw before 2026.4.10 contains an insufficient environment variable denylist vulnerability in its exec environment policy. Attackers can manipulate these environment variables to influence execution behavior and network connectivity.
Original NVD description (English source)
OpenClaw before 2026.6.5 could forward Authorization headers during MCP SSE redirects. When the affected feature is enabled and reachable, a lower-trust caller or configured input path could execute or persist actions beyond the caller's intended authorization. Impact depends on the operator's configuration and whether lower-trust input can reach the affected path.