发现优质的 AI Agent 技能

**Status:** Ready for publication **Version:** 1.0.0 **License:** MIT

**Date:** 2026-03-03 **Status:** ✅ Ready for publication **Version:** 1.0.0 **License:** MIT

来自 openclaw/skills 技能

`clawsaver`

This document shows how to integrate ClawSaver (session-level message batching) into OpenClaw's session handler.

**Status:** ✅ Complete & tested **Version:** 1.0.0 **Size:** 92 KB (4.2 KB core code + 87 KB docs/tests) **Tests:** 10/10 passing **Ready:** Yes, deploy anytime

Users sometimes send related messages in quick succession ("What is X? Give an example?"). Without batching, each message triggers a separate model call, doubling costs.

Print this and check off as you go.

Getting started with ClawSaver takes 10 minutes. Here's how to wire it in with minimal friction.

Use this skill to call the ClawShop backend API.

Audit a local OpenClaw install for security posture and common prompt-injection indicators. Produces a JSON report for review and alerting.

Human-in-the-loop security layer for OpenClaw. ClawShell intercepts shell commands before execution, analyzes their risk level, and requires your explicit approval (via push notification) for dangerous operations.