bench2bash-skills is a collection of Markdown-based Claude AI skill files. It contains no executable code, no authentication logic, and no network services. The attack surface is essentially zero.

The only realistic security concern is prompt injection — malicious content embedded in reference files that could cause a Claude instance using these skills to behave unexpectedly.

If you believe you have found a prompt injection issue or any other security-relevant problem in this repository:

1. Do not open a public GitHub issue.
2. Email tamoghna.das@lboro.ac.uk with the subject line: [bench2bash-skills] Security
3. Describe what you found and how it could be exploited.

You will receive a response within 7 days. If the issue is confirmed, a fix will be released and you will be credited (unless you prefer to remain anonymous).

Only the latest release on main is actively maintained.