agentraas

agentraas

Safety layer for AI agent actions
#102201320 followers5 following
⚡️ 4 day streak
#17,101Last month
0 KP
>10,000All time
0 KP

About

AgentRaaS is built by Sumedh, on his own. It sits between AI agents and the APIs they call: a retried call never runs twice, risky actions wait for a person to approve them in Slack, Teams, Telegram, Discord or email, spend caps stop runaway loops, and every action is logged. Open source and free to self-host on Linux, macOS or Windows. https://agentraas.io

Badges

Gone streaking
Gone streaking

Maker History

  • AgentRaaS
    AgentRaaSA safety layer for what your AI agents actually do
    • 🎉
    Joined Product HuntSeptember 8th, 2026

Forums

•

2d ago

AgentRaaS - Approvals, spend caps and undo for AI agent actions

New since the first launch: self-hosted Team features are free with no license key (approvals, SSO, spend-cap holds). Approve risky actions from Slack, Teams, Telegram, Discord, email or the dashboard. Daily money caps and allowed hours per agent, undo for refunds and messages, a one-command install on Linux, macOS and Windows, `agentraas scan` to check an MCP server before you connect it, and keys that can stay in your own Vault or cloud KMS.
•

2mo ago

AgentRaaS - Exactly-once execution for AI agents and automations

AgentRaaS sits between your AI agent (or n8n/Make/Zapier workflow) and the APIs it calls — Stripe, Twilio, HubSpot, or anything custom. It hashes and dedupes every call atomically, retries transient failures with backoff, trips a circuit breaker on real outages, and queues failures for one-click replay. Self-hosted is free and unlimited forever; hosted plans start at $20/mo.
View more