Skip to content

Skills

All management of agento-patronum happens via skills — the open specification for agent capabilities. Skills are invoked as slash commands inside Claude Code.

SkillDescription
/patronum-addAdd a pattern to the protection list
/patronum-removeRemove a pattern from the protection list
/patronum-listShow all protected patterns
/patronum-suggestGet stack-specific protection suggestions
/patronum-verifyRun self-test to verify hook enforcement

How they work

Each skill is a SKILL.md file following the agentskills.io specification. Skills instruct Claude to call the corresponding Node.js script via the Bash tool. The scripts read and write ~/.claude/patronum/patronum.json.

No external binary, no API calls — everything runs locally via Node.js.