Skip to content

Security: ToolsEnabled/toolsenabled-engine

Security

Report a vulnerability

SECURITY.md

Security policy

Reporting a vulnerability

Please do not report security problems in public issues. Report them privately in either of these ways:

Please include:

  • what you found and what it lets someone do;
  • the steps to reproduce it;
  • the release tag or commit, and your OpenShell version if it matters.

What to expect

  • We acknowledge a report within 3 business days.
  • We keep you informed while we work on it, and we credit you in the release notes unless you ask us not to.
  • Please give us a reasonable time to fix a problem before you disclose it publicly. We will agree a date with you.
  • There is no bug bounty.

What we ask of you

  • Follow this policy and any other relevant agreements. If this policy and other applicable terms disagree, this policy prevails.
  • Report a vulnerability promptly once you find it.
  • Avoid violating the privacy of others, disrupting our systems, destroying data or harming anyone's use of the product.
  • Discuss vulnerability information with us only through the two reporting channels above.
  • Give us a reasonable amount of time to resolve the issue before you disclose it publicly.
  • Test only what is in scope (below), and respect everything that is out of scope.
  • If a vulnerability gives you unintended access to data, access only the minimum needed to demonstrate it. Stop testing and report immediately if you come across anyone's personal information, health information, payment card data or proprietary information.
  • Only use accounts you own, or accounts whose holder has explicitly allowed you to use them.
  • Do not engage in extortion.

Safe harbor

When conducting vulnerability research, according to this policy, we consider this research conducted under this policy to be:

  • Authorized concerning any applicable anti-hacking laws, and we will not initiate or support legal action against you for accidental, good-faith violations of this policy;
  • Authorized concerning any relevant anti-circumvention laws, and we will not bring a claim against you for circumvention of technology controls;
  • Exempt from restrictions in our Terms of Service (TOS) and/or Acceptable Usage Policy (AUP) that would interfere with conducting security research, and we waive those restrictions on a limited basis; and
  • Lawful, helpful to the overall security of the Internet, and conducted in good faith.

You are expected, as always, to comply with all applicable laws. If legal action is initiated by a third party against you and you have complied with this policy, we will take steps to make it known that your actions were conducted in compliance with this policy.

If at any time you have concerns or are uncertain whether your security research is consistent with this policy, please submit a report through one of the reporting channels above before going any further.

Note that the Safe Harbor applies only to legal claims under the control of ToolsEnabled, Inc., and that the policy does not bind independent third parties. In particular, it does not cover research on NVIDIA's, Anthropic's or OpenAI's systems; their own policies apply there.

The "What we ask of you" and "Safe harbor" sections are adapted from the disclose.io policy templates (CC0 1.0).

Scope

In scope: the ToolsEnabled engine in this repository, including adapters/openshell/ (the image recipe, codex-openshell-auth, the sandbox policy) and the toolsenabled command (also installed as toolsenabled-openshell).

Report these to their owners instead:

If you are not sure where a problem belongs, send it to us and we will pass it on with your permission.

The ToolsEnabled desktop app is paused. It gets no new releases or fixes while paused. You can still report problems in it, and we will read them.

Supported versions

Only the latest published release of ToolsEnabled for OpenShell gets security fixes.

Where the security boundary is

In ToolsEnabled for OpenShell, OpenShell is the security boundary. It enforces the filesystem and network policy, holds the credentials it manages, and decides access approvals. ToolsEnabled's own rules (the work record, actor labels, the tool list) run inside the sandbox as the same user as the agents, so they are not a security boundary. The "Who enforces what" table in adapters/openshell/README.md lists each part.

No independent security review of ToolsEnabled has taken place yet.

There aren't any published security advisories