How-Tos & Practical Guides

The agent loop is only 1% of the problem. Databricks just announced what they’re doing about the other 99%. At the Data + AI Summit 2026 (DAIS 2026, June 15–18 in San Francisco), Databricks unveiled a major expansion of Agent Bricks — transforming it from a tool for building AI agents into a comprehensive developer platform for running them in production at enterprise scale. The numbers behind the announcement are striking: since Agent Bricks launched at last year’s DAIS, over 100,000 agents have been built on the platform, and Databricks is now processing 1+ quadrillion tokens per year in agent workloads. Customers like AstraZeneca, 7-Eleven, Fox Corporation, and Block have shipped production agents on Agent Bricks. ...

If you’re running agentic workflows in production, token costs are probably already on your radar. GitHub’s engineering team just published the most concrete, actionable deep-dive on agentic token optimization I’ve seen — and the results are striking: up to 62% reduction in token costs on their most-optimized workflows. This isn’t a theoretical exercise. These are lessons from the GitHub Agentic Workflows team — the people who maintain agentic automation that runs on every pull request across some of the most active repositories on the platform. Their findings are directly replicable in any agentic stack. ...

OpenClaw shipped version 2026.6.8 on June 16, 2026, and this is one of those releases where the quality-of-life improvements add up to something materially better. This isn’t a single-headline release. It’s a dense batch of improvements across channel delivery, model support, agent reliability, and UI polish — the kind of release that makes day-to-day agentic workflows feel less friction-heavy. Here’s what’s actually in it. Richer Telegram and WhatsApp Channel Delivery The headline change is structured message delivery for Telegram and WhatsApp channels. Previously, messages to these channels could lose formatting context — tables, nested lists, or rich structured text that rendered fine in a web UI might arrive as flat, hard-to-parse walls of text on mobile. ...

Anthropic Updates Privacy Policy — Claude Users May Face Mandatory Identity Verification (Gov ID + Selfie) From July 8 Anthropic revised its privacy policy around June 8, 2026, with the changes formally taking effect on July 8, 2026. The update introduces a new category of data collection called “Verification Data” and explicitly reserves the right to require identity verification from consumer Claude users. The process involves a government-issued photo ID and a live selfie. ...

AWS Summit New York 2026: Kiro, AgentCore, and Amazon Quick Are AWS’s Big Bets on Agentic AI The Javits Center in New York City hosted AWS Summit 2026 on June 17, and the keynote by Dr. Swami Sivasubramanian — VP of Agentic AI at AWS — made one message unmistakably clear: Amazon is going all-in on the infrastructure layer for production agentic AI. Three products dominated the keynote: Kiro, a spec-driven agentic IDE; Amazon Bedrock AgentCore, an enterprise platform for running agents at scale; and Amazon Quick, a reimagined successor to Amazon Q Business. Together they represent AWS’s answer to the question every enterprise is now asking: how do you go from agentic AI prototype to production system? ...

GitHub’s AI Agent Crisis Forces Microsoft to Tap AWS as Outages Break Enterprise SLAs Something remarkable and deeply uncomfortable is happening to GitHub: the platform built on the idea of distributed version control is now so centralized — and so overwhelmed by AI agent traffic — that Microsoft has had to quietly tap AWS to keep it running. A Microsoft-owned product, routing traffic through Amazon’s cloud, because the AI-generated commit volume has outpaced what Azure can absorb. ...

Hermes Agent Adds Asynchronous Subagents, So Delegated Work No Longer Blocks the Parent Chat One of the most persistent friction points in multi-agent orchestration has been the blocking nature of subagent delegation. You kick off a background task — scrape a dataset, analyze a codebase, run a long search — and your parent agent freezes, waiting for results that could take minutes. That’s over for Hermes Agent users. The Hermes Agent project just shipped asynchronous subagent support via a new async_delegation toolset. Delegated work now runs independently in the background while the parent session continues. This is a meaningful architectural shift that unlocks genuinely parallel multi-agent workflows. ...

OpenClaw 2026.6.8-beta.2 is out — and this release is a substantial quality-of-life upgrade for anyone running OpenClaw on Telegram, WhatsApp, or deploying agents with complex recovery needs. The 2026.6.x series has been moving fast since 2026.5.16-beta.4, and this beta brings together fixes and features across several areas: channel delivery, model support, usage reporting, and agent/gateway recovery. Here’s what’s new, what changed, and how to upgrade. What’s New in 2026.6.8-beta.2 Richer Telegram Delivery The biggest quality-of-life improvement in this release is Telegram’s new structured rich text support. OpenClaw agents can now deliver properly formatted tables, ordered and unordered lists, and expandable blockquotes through Telegram — which was previously reduced to plain text or awkward HTML approximations. ...

Salesforce Agentforce multi-agent orchestration reached General Availability on June 15, 2026 as part of the Summer ‘26 release — and it comes with a key architectural insight that any enterprise deploying it needs to understand: your agents’ natural-language descriptions are now the primary routing mechanism, not fixed decision trees. This changes what it takes to build reliable Agentforce implementations, and it means that teams who thought “configuration” was mostly technical setup are going to need to invest heavily in writing. ...

xAI shipped the Grok Build Agent Dashboard on June 15, 2026 — and if you’ve been trying to juggle multiple parallel coding agent sessions, this is the tool you’ve been waiting for. Instead of switching between terminal tabs and losing context on what each agent is doing, you now get a single terminal screen that shows all your active Grok Build sessions simultaneously. Here’s what it does, how to use it, and why parallel coding agent workflows are finally becoming practical for everyday development. ...