Phinq is an open-source runtime governance layer designed specifically for AI agents, built with the goal to stop AI agents before they make breaking or destructive errors.
As AI agents gain autonomy to execute tool calls, governance becomes critical. Phinq acts as a protective layer by intercepting every tool call made by an agent in real time. It analyzes and classifies each action based on risk level: safe, benign actions are automatically permitted to pass, while high-risk or irreversible actions are paused and held for human approval.
In addition to real-time risk classification and human-in-the-loop control, Phinq maintains complete accountability by recording every single decision into a tamper-evident, hash-chained audit log. This makes it a powerful developer tool for building reliable, safe, and auditable AI agent workflows.
Phinq is an open-source runtime governance layer for AI agents that intercepts tool calls to prevent agents from taking dangerous or breaking actions.
Phinq intercepts every tool call made by an AI agent and classifies it by risk. Safe actions are allowed to execute, whereas irreversible or risky actions are held until a human approves them.
Yes, Phinq records every decision and intercepted action into a tamper-evident, hash-chained audit log.
Yes, Phinq is open source and hosted on GitHub.
0 community reviews for Phinq
Help the community by sharing your experience. Your review helps others make better decisions.
I build things.
