Skill Auto-Use

Automatically use installed skills without being asked. Maintain a trigger table that maps contexts to skills, and enforce that every newly installed skill g...

This skill is a policy that tells your agent to automatically run any installed skill that matches a trigger and to persist trigger rules in workspace files. It doesn't install code or ask for credentials itself, but it hands the agent broad authority to call other skills (some of which may contact external services or access sensitive data) without asking you. Before installing, consider: (1) only enable auto-use for non-sensitive skills, (2) maintain an explicit allowlist/denylist (do not auto-add skills that use external APIs or have write/destructive actions), (3) require explicit user consent for any skill that accesses external services or credentials, (4) avoid making this policy permanent in shared workspaces, and (5) test in a safe environment to ensure it doesn't cause loops or mass invocations. If you want to retain manual control, prefer leaving the skill uninstalled or set the agent policy so skills are only used after explicit user confirmation (or set disable-model-invocation to true).

Type: OpenClaw Skill Name: skill-auto-use Version: 1.0.0 The skill 'skill-auto-use' (SKILL.md) instructs the AI agent to bypass user confirmation and automatically execute installed skills based on context triggers. It mandates the creation of a 'trigger table' and uses strong imperative language ('No exceptions', 'Don't ask... Just use it') to override default safety behaviors. While intended for automation, this behavior significantly increases the risk of unauthorized or unintended tool execution by removing human-in-the-loop oversight.