Facio Blog
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.
Showing 66-70 of 81 articles.
Human-in-the-loop
Not every agent action needs human review — and not every action can run autonomously. A practical risk-tiered decision framework for HITL approval workflows.
Engineering
Mockzilla lets AI agents spin up API mocks from any OpenAPI spec in seconds. 2,200+ specs validated, 98,000+ endpoints — here's how to integrate it with Facio for agent-driven API testing.
Security
73% of production AI agent deployments are vulnerable to prompt injection. A practical guide to building layered defenses that actually work in production — from input sanitization to HITL gates.
Engineering
WhoDB CLI brings PostgreSQL, MySQL, SQLite, MongoDB, and Redis to your AI agent via MCP. Natural language queries, schema visualization, and read-only safety — here's how to integrate it with Facio.
Security
The OWASP Agentic Top 10 defines the threat model for AI agents. Goal hijack, tool misuse, MCP supply chain — here's what each risk means and how to harden your agent stack.