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.

Showing 161-165 of 235 articles.

Security

Your Monitoring Says Green, Your Agent Is Wrong: The Observability Gap Killing Production AI

Jun 8, 2026Security

Your Monitoring Says Green, Your Agent Is Wrong: The Observability Gap Killing Production AI

Traditional monitoring can keep 99.99% availability while the agent's decisions degrade quality across the workflow. Decision integrity is the new observability surface — and OpenTelemetry's GenAI semantic conventions plus the new MCP tracing layer are finally making it tractable. Here's how to build for it.

Engineering

MCP Spotlight: ComparEdge — Give Your Agent Instant SaaS Pricing, Alternatives & Side-by-Side Comparisons

Jun 7, 2026Engineering

MCP Spotlight: ComparEdge — Give Your Agent Instant SaaS Pricing, Alternatives & Side-by-Side Comparisons

ComparEdge gives AI agents 9 MCP tools to search, compare, and price 508+ SaaS and AI products across 45 categories — with verified vendor pricing, G2/Capterra ratings, and per-token LLM costs. Zero config, no API key required.

Product

Facio's Media Generation Tools: How AI Agents Create Images and Videos Programmatically

Jun 7, 2026Product

Facio's Media Generation Tools: How AI Agents Create Images and Videos Programmatically

AI agents that can only produce text are leaving half their potential on the table. Facio's generate_image and generate_video tools let agents create visual content programmatically — across OpenAI, Google Gemini, Replicate, and fal.ai — with provider-agnostic APIs, HITL approval gates, and direct delivery to any channel. Here's how autonomous visual content creation works in production.

Security

The Sentence That Stole an npm Token: How AI Agents Became the Third Class of Lateral Movement

Jun 7, 2026Security

The Sentence That Stole an npm Token: How AI Agents Became the Third Class of Lateral Movement

A single GitHub issue title — just a sentence — compromised a CI/CD pipeline and published a poisoned npm package. No network intrusion. No credential theft. Just an AI agent bridging an untrusted comment field and a privileged software supply chain. Welcome to agent-mediated lateral movement.

Human-in-the-loop

When NOT to Use Human-in-the-Loop: The Autonomy-First Decision Framework

Jun 7, 2026Human-in-the-loop

When NOT to Use Human-in-the-Loop: The Autonomy-First Decision Framework

HITL is not always the right answer. For some workflows, it adds cost without value — latency without safety improvement. Here's a decision framework for identifying when human oversight should be removed, reduced, or replaced with automated guardrails.