CVE-2026-83816codewhale-hq CodeWhale OS command injection
CVE-2026-83816 is a OS command injection in codewhale-hq CodeWhale, rated high (8.1 on CVSS 3.1).
Why this priority: High severity, with no exploitation signal yet.
Published · Updated · NVD status: Received
What it is
Input reaches an operating-system command line, letting an attacker run their own commands on the host. A permission check exists but gets the answer wrong for some callers.
How it is reached
An attacker can exploit it over the network, without an account, and nobody has to click anything. Success also depends on conditions the attacker does not fully control.
What an attacker gains
A successful attack can read the data it protects, change data or behaviour and take it offline. In practice, this kind of flaw lets an attacker run commands or code with the privileges of the affected service.
Is it being exploited?
There is no public report of exploitation yet.
What to do
- Upgrade codewhale-hq CodeWhale to 0.9.0 or later.
- Until it is patched, keep it off the internet or behind a VPN, allowlist or authenticating proxy.
Description from NVD
CodeWhale is a DeepSeek and MiMo coding agent for the terminal. Prior to 0.9.0, the ExecShellTool::approval_requirement_for path used is_parallel_readonly_command to downgrade commands to automatic approval based on a flag-stripped command allowlist. The allowlist treated fd, rg, and git grep as read-only even when fd -x or fd --exec, rg --pre, or git grep --open-files-in-pager selected an external program. registered_tool_approval_required returned without displaying a prompt for that automatic classification, and the later command-safety check also treated the base utilities as safe. Attacker-controlled content that influenced an agent tool call could therefore cause arbitrary program execution on the developer host without the approval promised by the default interactive profile, resulting in unauthorized access, modification, or disruption. This issue is fixed in version 0.9.0.
What this means for your architecture.
A CVE names a product. Whether it matters depends on where that product sits in your systems: what reaches it, which trust boundary it guards and what it can reach once compromised. Veriom maps your components, their relationships and your trust boundaries from your code, cloud and delivery evidence, so you can see where a component like this sits. The radar itself does not look at your systems.
Components it touches
- Application servers and the hosts they run on. The flaw sits between untrusted input and code or command execution.
- Template engines. The flaw sits between untrusted input and code or command execution.
- Permission checks in APIs and admin surfaces. The flaw sits between one role or tenant and another.
- Multi-tenant services. The flaw sits between one role or tenant and another.
Trust boundaries in play
- Internet-facing entry points. No account is needed. Any instance reachable from outside your perimeter, through a load balancer, gateway, VPN or exposed admin port, is in scope.
- Between untrusted input and code or command execution. An attacker who crosses it can run commands or code with the privileges of the affected service.
- Between one role or tenant and another. An attacker who crosses it can reach data or actions their role should not allow.
Questions to answer about your own systems
- Where does codewhale-hq CodeWhale run in your systems, and at which version?
- Which of those instances can be reached from the internet or partner networks?
- What can it reach once compromised: which data stores, secrets and services?
Sources
- CVE records and CVSS scores from the NVD CVE API. This product uses the NVD API but is not endorsed or certified by the NVD.
- Exploitation status from CISA’s Known Exploited Vulnerabilities catalog.
- Exploit likelihood from FIRST’s Exploit Prediction Scoring System (EPSS), refreshed daily.
- Explanations are written by Veriom from fixed templates over each record’s data, not by a person reviewing each CVE. Descriptions are quoted from the records as published.