Security fixes are made on main and ship in the next release. Only the latest release is supported; please reproduce on it (or on main) before reporting.

Please do not open a public issue, pull request or discussion for a security problem.

Report it privately through GitHub instead: Report a vulnerability (Security tab → Report a vulnerability).

A useful report includes:

• the Mulder version or commit, and how it was run (Docker image, install.sh, local uv install) and with which model provider;
• the affected component (for example a tool wrapper, the MCP server, the audit log, case handling or report generation);
• steps to reproduce, ideally with a minimal evidence file or directory;
• the impact you expect, and a suggested fix if you have one.

You will get a reply in the advisory thread. Please keep the details private until a fix is released and the advisory is published.

Mulder processes evidence that is attacker-controlled by design, so the following are in scope:

• crafted evidence (disk images, memory dumps, PCAPs, logs, archives) that writes, reads or executes outside the case directory, runs commands, or crashes or hangs an investigation;
• content in evidence that bypasses a security boundary, for example forging or altering audit-log entries or evidence citations, or making a role or tool act outside its permissions;
• injection into generated reports (HTML, IOC exports) or into tool command lines;
• leaks of credentials or evidence content to places they should not go.

Out of scope:

• vulnerabilities in third-party forensic tools themselves (please report those upstream; issues in how Mulder invokes or isolates them are in scope);
• wrong or missed findings that do not cross a security boundary - please open a regular issue for those.