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 56-60 of 235 articles.

Engineering

MCP Spotlight: Linear MCP Server — The Official Issue-Tracking Bridge With Initiatives, Project Milestones, Code Intelligence, and the Engineering-Workflow Default for Agents

Jul 9, 2026Engineering

MCP Spotlight: Linear MCP Server — The Official Issue-Tracking Bridge With Initiatives, Project Milestones, Code Intelligence, and the Engineering-Workflow Default for Agents

The official Linear MCP Server — ~40 focused tools covering issues, projects, cycles, initiatives, project milestones, updates, comments, customers. February 2026 added Initiatives + Project Milestones for parity with the web app. MIT-licensed, OAuth 2.0 + Personal API Key support. The issue-tracking + project-management default for engineering agents in 2026.

Product

Facio's Distributed Tracing Discipline: How AI Agents Stay Debuggable When Every Step Crosses a System Boundary

Jul 9, 2026Product

Facio's Distributed Tracing Discipline: How AI Agents Stay Debuggable When Every Step Crosses a System Boundary

An AI agent's execution is a chain of model calls, tool invocations, database queries, API requests, and human-in-the-loop pauses. When something goes wrong — the agent makes a wrong decision, the workflow takes too long, the output is wrong — the team needs to trace the failure to its root cause across every system involved. Facio's distributed tracing discipline gives teams the structural observability to debug AI agents the way they debug microservices: trace IDs propagated through every step, spans for every tool call, and correlated logs across all systems.

Human-in-the-loop

HITL and the Cost of Friction: Why Your Reviewers Are Quitting and How to Architect Against It

Jul 9, 2026Human-in-the-loop

HITL and the Cost of Friction: Why Your Reviewers Are Quitting and How to Architect Against It

Every HITL interface adds friction — structured reasoning fields, minimum time enforcement, mandatory doubt capture, two-reviewer requirements. The friction is justified by the quality it produces. But friction has a cost: the reviewers burn out, the best ones quit, the worst ones stay. Here's how to design HITL friction that improves quality without destroying the reviewer pool.

Engineering

MCP Spotlight: Cloudflare MCP Server — The Two-Tool `search()` + `execute()` Bridge to the Entire 2,500-Endpoint Cloudflare Platform

Jul 8, 2026Engineering

MCP Spotlight: Cloudflare MCP Server — The Two-Tool `search()` + `execute()` Bridge to the Entire 2,500-Endpoint Cloudflare Platform

The official Cloudflare MCP Server — the boldest tool-surface design in the MCP ecosystem. Just 2 tools (search + execute) expose the entire 2,500+ endpoint Cloudflare API (Workers, R2, D1, DNS, AI Gateway, Vectorize, Zero Trust, Stream). MIT-licensed, official Cloudflare-maintained. The minimal-surface design pattern at its most extreme.

Product

Facio's Kill Switch: How AI Agent Workflows Stop the Moment Something Goes Wrong

Jul 8, 2026Product

Facio's Kill Switch: How AI Agent Workflows Stop the Moment Something Goes Wrong

Every production AI agent workflow needs a kill switch — a way to stop all in-flight executions the moment something goes wrong. The agent that processed 1,000 customer refunds correctly before hitting an edge case that causes incorrect refunds; the agent that compiled 50 PRs cleanly before encountering a malicious dependency; the agent that published 100 social posts before being tricked into a policy violation. Without a kill switch, the damage compounds. With one, damage stops at the moment of detection. Facio's kill switch discipline is what makes autonomous AI agent operations survivable.