excessive-agency

What problem does it solve?

Prevents autonomous agents from taking irreversible or high-impact actions without human oversight, reducing risk when deploying agentic workflows that interact with files, emails, databases, and external services.

Core Features & Use Cases

• Action gating by impact — high-impact actions require human approval before dispatch.
• Tool boundaries and allowlists — restricts actions to defined, safe resources and endpoints.
• Audit-logging and traceability — records decision context, approvals, and actions before dispatch.
• Safe tool design guidance — promotes structured interfaces to prevent arbitrary code or commands.
• Use Case: When an LLM plans to modify production data or send external communications, safeguards ensure oversight and accountability.

Quick Start

Configure an agent workflow to require human approval before executing high-impact actions and log every action before dispatch.