CVE-2026-53853: OpenClaw: Linux and macOS exec allowlists skipped configured argument patterns

Exec allowlist entries can carry an optional argPattern that restricts the arguments for an executable. With tools.exec.security set to allowlist, the Linux and macOS gateways treated a matching executable path as enough and never evaluated the pattern. The tool description still advertised it as enforced. Windows was not affected, because its code path already checked argPattern. Path-only entries behave as intended, since they allow any arguments by design.

The attacker needs to influence a tool-enabled agent to call exec, for example as a lower-trust sender on a channel the agent reads, or through prompt injection. They then run arguments the operator meant to forbid. What that gives depends on the executable. The advisory names git, python, node, bash, find, tar and ssh as high risk, because interpreter, loader, subprocess, network or plugin flags can mean file access, network access or command execution that should have asked for approval. The advisory limits the claim: it is a bypass of the direct argPattern predicate, not of all exec approval semantics.

The npm package openclaw before 2026.5.12, when all of these hold: the gateway runs on Linux or macOS, tools.exec.security is "allowlist", at least one allowlist entry uses argPattern, and that executable accepts security-relevant flags. Check:

npm ls -g openclaw

Then review your exec allowlist for entries that combine a path with argPattern.

Upgrade to openclaw 2026.5.12, the first stable patched version.

npm install -g [email protected]

Before upgrading, the advisory tells operators in allowlist mode to review entries that pair an executable path with argPattern, especially for interpreter-like or subprocess-capable tools. Treat those entries as unrestricted until patched, and remove the ones you cannot justify.

List your allowlist entries that use argPattern. Then review exec history for those executables, looking for arguments that your pattern would have rejected, and for runs that did not raise an approval prompt. Pay attention to interpreter flags such as -c or -e, ssh options, and find -exec. Beam records the shell commands an agent runs and flags risky ones, so off-pattern invocations of an allowlisted binary can be found there.

Summary

OpenClaw's exec allowlist supported optional argPattern entries to restrict the arguments accepted for an allowlisted executable. In affected releases, Linux and macOS gateways skipped argPattern checks and treated a matching executable path as sufficient to satisfy the allowlist.

This meant an operator could configure an allowlist entry that appeared to permit only a narrow argv shape, but OpenClaw would allow other argv for the same executable without an approval prompt when tools.exec.security was set to allowlist.

This issue is limited to direct enforcement of configured argPattern values. OpenClaw's exec approvals remain best-effort guardrails and do not attempt to semantically model every interpreter, loader, package script, shell feature, or transitive file a command may use.

Affected configurations

This affects OpenClaw gateway deployments that meet all of these conditions:

• the gateway runs on Linux or macOS
• exec is configured with tools.exec.security: "allowlist"
• at least one exec allowlist entry uses argPattern
• the allowlisted executable accepts security-relevant arguments or flags

Path-only allowlist entries are not additionally affected by this issue, because those entries intentionally allow any arguments for the matched executable. Windows was not affected by this specific bug because the affected code path already applied argPattern checks on Windows.

Impact

If an untrusted or lower-trust sender can influence a tool-enabled agent to call exec, they may be able to run disallowed arguments for an executable that the operator intended to restrict with argPattern. Depending on the executable, those arguments can cause host-side file access, network access, or command execution that should have required an approval prompt.

The practical impact depends on the operator's allowlist and channel exposure. Examples of higher-risk allowlisted executables include tools with interpreter, loader, subprocess, network, or plugin flags such as git, python, node, bash, find, tar, and ssh.

This is not a bypass of all exec approval semantics. It is a bypass of the direct argPattern predicate that the operator configured and that the exec tool description advertised as enforced at runtime.

Patched Versions

The first stable patched version is 2026.5.12.

Mitigations

Upgrade to [email protected] or later. Before upgrading, operators who use exec allowlist mode should review entries that combine an executable path with argPattern, especially for interpreter-like or subprocess-capable tools.