Use GitHub's private vulnerability reporting (Security → Report a vulnerability on this repository). Please do not open a public issue for anything exploitable.

Include: affected paths/scripts, reproduction steps, and the environment (OS, Python version, harness).

• The kit runs your code through your agent harness; prompt-injection resistance is a methodology goal (see eval/scenarios/), not a sandbox. Do not point agents at untrusted repositories and expect isolation.
• scripts/install.py writes only under MEMORY_ROOT (default ~/.memory) and links the engine from the clone — review it before running, like any bootstrap script.