AgenRACI defines who is accountable when an AI agent acts

AgenRACI is an open-source tool for writing down who is accountable when an AI agent ships code, replies to a customer, or spends money. Teams write one file that lists the actor, accountable owner, approval path, permissions, timeout, and emergency behavior for each type of action. Its checker finds structure problems such as missing owners, two accountable owners, unused permissions, approval paths with no timeout, and escalation loops. It does not block tool calls or enforce approvals at runtime.

Key points

  • AgenRACI uses one file to define responsibility rules for agent actions.
  • The checker can fail in CI when the responsibility rules have gaps.
  • It covers approval paths, timeouts, permissions, and emergency behavior.
  • It is independent of agent frameworks such as LangGraph and CrewAI.
  • It writes and checks the charter, but does not enforce approvals at runtime.

Quick term guide

open-source
Software whose code is shared publicly so others can inspect, use, or change it.
permissions
Settings that define what files or actions a system or user is allowed to access.
escalation
When an AI or lower-level support agent passes a problem to a human or higher-level support because it cannot solve it.
tool calls
Times when an AI system uses another function, such as search or file access.
governance
The policies and controls a company uses to manage data and systems safely and in compliance with rules.
model calls
Requests sent to an AI model to get an answer or action.
agent frameworks
Developer toolkits that help AI models use multiple tools in sequence to complete complex tasks automatically
frameworks
Pre-built templates and tools that make making websites easier.
Read original