Agents

AI Agent API

Connect AI agents (OpenClaw, LangChain, CrewAI, GPT Actions) to Webhook Hub. Agents can discover the API, listen for events with suggested actions, and take automated corrective action — create tickets, set up alerts, toggle forwarding, replay events.

Endpoints

API Discovery

GET

Machine-readable OpenAPI spec. Point any agent framework at this URL to auto-discover all endpoints.

Event Feed

GET

Recent events formatted for agent consumption. Each event includes suggested actions and available API actions the agent can execute.

Agent Actions

POST

Execute actions programmatically. Agents can create rules, replay events, toggle forwarding, and more.

  • replay_event — re-process from raw payload
  • create_forwarding_rule — set up email/Slack/SMS/call/webhook
  • create_playbook — attach remediation steps to events
  • create_automation — trigger action chains on events
  • create_alert_rule — monitor metrics with thresholds
  • toggle_forwarding_rule — enable/disable a rule

AI Analysis

POST

Analyze recent events with Claude AI. Returns summary, patterns, risks, and actionable recommendations. Falls back to structured analysis without API key.

Quick Start

Copy this into your agent framework to get started:

# 1. Discover the API curl # 2. Get events with suggested actions curl # 3. Take an action curl -X POST \ -H "Content-Type: application/json" \ -d '{"tenant_id":"demo_tenant","action":"create_alert_rule","params":{"name":"Agent alert","metric":"error_rate","threshold":25}}'

Test It

Click to fetch the agent feed for demo_tenant and see what an agent receives: