adversarial-boundary-gate

name	adversarial-boundary-gate
description	Pre-delivery threat-model of trust boundaries and an abuse-case/injection check, delegating to a fresh-context security reviewer.
metadata	{"short-description":"Pre-delivery threat-model of trust boundaries and an abuse-case/injection check, delegating to a fresh-context security reviewer."}

adversarial-boundary-gate

This is a thin adapter for agents that cannot load symlinked skills.

Canonical skill source:

• ~/ai-agents-skills/canonical/skills/adversarial-boundary-gate/SKILL.md

Before using this skill, read the canonical source file above and follow its instructions. Related reference files live next to that source file in the same skill directory.