Agent Skills
One SKILL.md, a trigger description and a few optional scripts turn a working method into a shareable capability. Skill libraries now circulate between Codex, Hermes and OpenHands, so check the source, the permissions and what the scripts do before installing.
A skill packages a working method into a directory: a YAML header carries the name and trigger condition, the body holds the steps, scripts load on demand. Codex scans .agents/skills upward, picks a skill by $.skill-name or from its description, and injects SKILL.md only once selected, so a vague description never fires. Sharing a capability shares the risk: third-party skills are screened for exfiltration and prompt injection.
