Saturday, November 7, 2026 270 picks · 133 builders credited Subscribe
AgentRags.

The agentic rag, wrung out.

The rag for

Friday, October 23, 2026

The day in one breath

Hermes Agent v0.22.5 modularization and OpenClaw Enterprise launch

Hermes moved core features (MCP server mode, import-agent, OpenClaw migration tooling, Singularity backend) into plugins while shipping per-task subagent routing and MCP spend controls. Simultaneously, OpenClaw publicly outlined its Enterprise isolation model (trusted control plane vs untrusted agent workloads) and opened pilot conversations. Together they mark a clear shift: both systems are hardening the boundary between core agent runtime and extensible/enterprise surface area. This is the highest-signal architectural story for builders right now.

Releases

Hermes Agent v0.22.5 Ships Major Architecture Shifts

v0.22.5 (stable, Oct 23) moves hermes mcp serve, import-agent, OpenClaw migration, and Singularity backend out of core into official plugins. Adds per-task subagent routing (model or profile via advanced-delegation plugin), usage-window gauges for Nous plans (/usage bars), MCP tool cost declarations with per-chat spend limits, Edge TTS fix, Pillow 13 security bumps, and Electron 44 desktop updates. Detailed 10-item breakdown covers before/after behavior.

Guides & Workflows

/usage Fuel Gauge for Nous Usage-Window Plans

On Nous plans that track daily/weekly/monthly allowances (instead of pure credits), /usage now shows one bar per limit with percent used and time to reset. Model picker tags Large vs Free. At limit, Hermes names what ran out, when it resets, and what still works — without auto-switching models. Desktop gauges coming with next app update.

Multi-agent & Memory

Hermes Memory Setup Walkthrough

Practical guide to `hermes memory setup`. Hermes keeps MEMORY.md / USER.md; a memory provider runs alongside for semantic search, knowledge graphs, and automatic fact extraction. One command walks through provider selection. Built-in memory stays active. Good for long-running projects or daily agents.

Releases

OpenClaw Enterprise Architecture and Isolation Model

Official thread on OpenClaw Enterprise (pre-1.0, pilots open). Separates trusted control plane (RBAC, audit logs, enforced policy) from untrusted agent/tool workloads in distinct isolation boundaries. Designed to plug into existing sandboxes, MCP gateways, and secret brokers via a common contract. Red Hat and NVIDIA contributions noted. Slack for enterprise discussion.

Never miss a night

Every pick, delivered to your reader.

One feed for everything, or one per agent harness you run. Rebuilt every night, with full credit on every item.