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 6-10 of 235 articles.

Product

Facio's Interpretability Discipline: How AI Agents Make Decisions That Customers, Auditors, and Regulators Can Actually Understand

Jul 23, 2026Product

Facio's Interpretability Discipline: How AI Agents Make Decisions That Customers, Auditors, and Regulators Can Actually Understand

AI agents make decisions that affect real people: a loan is approved or denied, a claim is paid or rejected, a customer is offered a discount or not, content is flagged or allowed. The people affected deserve an explanation. Auditors demand one. Regulators require one. Naive agents produce outputs without explanations; the team can't tell why the agent decided what it decided. Facio's interpretability discipline gives every decision a structured explanation: what data the agent saw, what reasoning it followed, what alternatives it considered, and why it chose this path.

Security

NIST AI RMF Agentic Profile Implementation Guide: The 60 Controls That Map Every Architecture Pillar to the Four Core Functions

Jul 23, 2026Security

NIST AI RMF Agentic Profile Implementation Guide: The 60 Controls That Map Every Architecture Pillar to the Four Core Functions

The NIST AI Agent Interoperability Profile is planned for Q4 2026 release. The 60-control implementation guide maps every architecture pillar from runtime policy engines to kill switches, behavioral analytics, supply chain security, and HITL workflows to the four AI RMF core functions (Govern, Map, Measure, Manage) and the EU AI Act, GDPR, NIST 800-53, and ISO 27001 frameworks.

Human-in-the-loop

HITL and the Second-Order Question: Why the First Action's Outcome Determines Whether the Next Review Is Even Possible

Jul 23, 2026Human-in-the-loop

HITL and the Second-Order Question: Why the First Action's Outcome Determines Whether the Next Review Is Even Possible

Most reviewers think about the action in front of them. Few think about whether the action enables the next review. The first-order question is "is this right?" The second-order question is "does approving this preserve my ability to review the next one?" That second-order question is what separates review systems that scale from review systems that collapse into theater. Here is why the second-order view is the hidden architecture of HITL.

Engineering

MCP Spotlight: Slack MCP Server — The Official Team-Communication Bridge With LLM-Native Tool Design, Channel-Add Allow-List, and the Conversation-Default Reference for Agents

Jul 22, 2026Engineering

MCP Spotlight: Slack MCP Server — The Official Team-Communication Bridge With LLM-Native Tool Design, Channel-Add Allow-List, and the Conversation-Default Reference for Agents

The official Slack MCP Server by Slack — ~30+ LLM-native tools (built for LLM consumption with rich descriptions, natural-language responses, examples, pre/post-call hints). Posting disabled by default, enabled per-channel via explicit allow-list. OAuth 2.0 with principle-of-least-privilege scopes. MIT-licensed.

Product

Facio's Anti-Abuse Discipline: How AI Agent Systems Detect and Stop Prompt Injection, Loops, and Exfiltration Before Damage Is Done

Jul 22, 2026Product

Facio's Anti-Abuse Discipline: How AI Agent Systems Detect and Stop Prompt Injection, Loops, and Exfiltration Before Damage Is Done

AI agents in production face a new category of abuse that traditional systems weren't built for. Prompt injection turns instructions into weapons. Cost-exhaustion loops burn through thousands of dollars in minutes. Data exfiltration smuggles customer data out through tool calls. Token-budget attacks exploit context windows. Without anti-abuse discipline, AI agents become unwitting accomplices. Facio's anti-abuse discipline gives the runtime the structural defenses: input sanitization, loop detection, exfiltration monitoring, and circuit-breaker isolation.