Agentops: Local to Cloud — wallets, tracing, and orchestration

Deploy local agents everywhere with LFM2.5-2.6B. LiquidAI releases a tiny agent model (LFM2.5-2.6B) that runs tool-using, multi-step agents on-device, letting phones and laptops host private agents. Outcome engineers can shift work off cloud GPUs into local, private deployments — rethink orchestration, sandboxing, and offline validation (Principles 07, 09).

Introducing: Cloudflare Agents. Cloudflare launches Agents and agent tracing to centralize deployment, observability, and token-level insights across hosted agent sessions. This gives outcome engineers the tracing and token accounting needed to make agent behavior legible and auditable in production (Principles 06, 16).

Announcing Cloudflare Wallets: the programmable wallet for the agentic Internet. Cloudflare ships programmable wallets and identity primitives so agents can authenticate and pay for APIs under human-set guardrails. Treat agent wallets as a new security and UX surface: design policy gates, spend controls, and consent flows into your agent workflows (Principles 03, 04, 15).

AWS’s Kiro Crew aims to turn AI coding agents into autonomous engineering teams. AWS open-sources Kiro Crew, an orchestration platform for persistent, autonomous coding agent teams that carry context across tools and repositories. Outcome engineers can adopt persistent-context and team-orchestration patterns but must instrument routing, cost controls, and human-in-the-loop checkpoints (Principles 09, 06).

How we built a software factory to drive Astro’s GitHub issue count to zero. Cloudflare describes Flue, a software-factory pattern that runs isolated agent subagents in GitHub Actions to automate triage and reduce issue backlog. Use this as a concrete blueprint: agents-as-CI lanes, isolated execution, and executable artifacts plus observability to ensure safe delivery (Principles 07, 09, 08).