Integrations

Agent trust layer · Works with your stack

LetsPing helps you secure your AI agents. You get a behavioral firewall and human-in-the-loop at the tool boundary: wrap dangerous tools, we intercept anomalies, ping you to approve, and learn. One SDK works with LangGraph, Vercel AI SDK, CrewAI, or any runner. Separately, we offer an agent-first layer: verifiable credentials, signed handoffs, and usage-based billing so agents can transact without a human. One platform whether you want human-in-the-loop or the agent economy.

• LangGraphAgent frameworkRecommended

  LetsPingCheckpointer, defer/approve, Cryo-Sleep state. Production HITL for LangGraph.

• Vercel AI SDKAgent framework

  Wrap tools with LetsPing; stream with approval gates and behavioral firewall.

• CrewAIAgent framework

  LetsPingCrewTool for human approval on dangerous actions. Python integration.

• SlackNotifications

  Approve or reject from Slack. Connect in Settings → Integrations.

• OpenClawDesktop / CLI

  HITL and human approval for OpenClaw. Terminal pairing; approve or patch from the LetsPing PWA.

• PythonLanguage

  letsping package on PyPI. LangGraph and CrewAI helpers.

• Node.jsLanguage

  @letsping/sdk on npm. TypeScript and JavaScript.

Full API reference · Agent APIs · Home