Facio Blog

Practical notes on human-reviewed AI agents.

Payload-powered product notes, security writing, HITL patterns, and operational guidance from the Facio runtime: long sessions, Placet approvals, audit trails, memory, providers, channels, tools, and Docker-first operations.

Clear

Showing 71-75 of 78 articles in Product.

Product

How Facio's Credential Store Prevents AI Agent Secret Leaks at Runtime

May 28, 2026Product

How Facio's Credential Store Prevents AI Agent Secret Leaks at Runtime

In 2025 alone, nearly 29 million new secrets were found hardcoded in public GitHub commits — and AI-assisted code leaks at double the rate. Facio's credential store solves this by ensuring agents never touch raw API keys. Here's how the architecture works, from password-gated setup to write-time placeholder resolution to full audit trail coverage.

Product

How Facio Enforces Human-in-the-Loop at the Runtime Level — Not in Agent Code

May 26, 2026Product

How Facio Enforces Human-in-the-Loop at the Runtime Level — Not in Agent Code

Most agent frameworks implement human-in-the-loop as a code-level interrupt. Facio enforces it at the runtime level — with structured approval gates, configurable timeouts, audit trails, and a dedicated human review interface. Here's how the architecture works, and why the governance layer belongs in the platform, not the prompt.