withHuman
Open-source human approval layer that intercepts agent tool calls before they run.
withHuman is an open-source approval layer that intercepts AI agent tool calls before they run. Rules can approve or deny routine calls automatically, and consequential actions go to the right person via Slack, email or mobile. It works across Claude Code, Codex, LangGraph, Pydantic AI and more through the open Agent Approval Protocol.
- Category
- Infrastructure
- Pricing
- Open Source
- Website
- withhuman.ai
Related tools and alternatives
E2B
e2b.devSecure cloud sandboxes for AI agents to run code.
Composio
composio.devTool integrations and auth for AI agents.
Navige
navige.aiCost-control layer that sits between AI agents and the models and tools they pay for.
Agent37
agent37.comLow-cost cloud sandboxes for hosting AI agent harnesses with persistent storage.
Claude Code
claude.comAnthropic's agentic coding tool that works in your terminal and codebase.
OpenAI Codex
openai.comOpenAI's cloud and CLI software engineering agent.