This is the skill you want when a user says "help me set a goal" or asks for the goal tool, and you need to turn fuzzy intention into something measurable before work starts. It walks you through restating vague requests as concrete objectives with verification evidence, quantitative thresholds, and scope boundaries. The quality bar is strict: no "make progress" or "investigate X" unless you can sharpen it into a pass/fail validator, a metric with a target, or specific evidence counts. It checks for active goals before creating duplicates and refuses to create goal artifacts for ordinary multi-step tasks unless explicitly requested. Think of it as a gatekeeper that forces honest objectives before spinning up goal-backed work.
npx -y skills add openai/skills --skill define-goal --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
sickn33/antigravity-awesome-skills
moizibnyousaf/ai-agent-skills
github/awesome-copilot