A human-approval gate for the dangerous things an AI agent does.

Pi-family policy gate between an AI agent and an onchain wallet: propose a payment, the gate allows it, holds it for human approval, or blocks it, with an append-only receipt log. Optional zero-knowledge spending policy (ERC-8366): a Groth16 proof that an

Logged per-agent plan mode with deployment guidance, a direct slash command, and a user-reviewed exit

Transactional guard proxy for MCP agents: impact diff, human approval gate, saga compensation, hash-chained audit log.

MCP server for the Presence Human Approval Gate: pause an AI agent's consequential tool call until a named person approves it, with a sealed Presence Record either way.

MCP proxy server that intercepts dangerous AI agent tool calls for human approval via Slack or email

MCP stdio proxy for governing tool calls with ActionProxy policy, human approval, one-time grants, and audit

TOTP-authenticated command gateway - intercept dangerous shell commands, require human approval

Paraph SDK — tamper-evident records and human-approval gates for AI agents. EU AI Act Articles 12 + 14.

Human approval for AI agents with a signed audit trail — JS/TS client for Raposa Aval.

Sandclaw Builder plugin — Docker-based code builds via Confidante with human approval

Enterprise approval workflow plugin for Agorio SDK — require human approval before high-value transactions

Verifiable spend-control policies for Hedera AI agents — drop-in Hedera Agent Kit hooks for budgets, allowlists, human approval, and an immutable HCS audit trail.

Security gateway for Model Context Protocol servers: policy enforcement, secret redaction, prompt-injection screening, tool rug-pull detection, rate limiting, human approval gates, and audit logging — one drop-in proxy, zero infrastructure.

SignedApproval n8n node — cryptographic human approval with Ed25519 proof

Local human approval server and page for PermitRail agent tool calls. For demos, single-user dev, and internal tools.

Turn a signed provider webhook into a **durable signal** a parked flow can `await`. This is the inbound edge that lets a durable run pause on an external event — a `/qa run` comment, a preview deploy, a payment, a human approval — and resume crash-safe wh

Anti-slop guardrails, specialized agents, and a live observability + control plane for AI-assisted coding. Watchtower (Python + SQLite, no setup) shows a live heartbeat, your sessions, and a needs-human approval queue — checkpoint, compact, fork, close, o

Human-friendly process signals

OKed for the Claude Agent SDK. A ready-made PreToolUse hook that gates sensitive tool calls behind a human approval push to your phone.