El blog de Workstation
AI Agents
Publicaciones etiquetadas con AI Agents.
Hybrid Execution for Coding Agents: Macro Skills, Micro Recovery
Bundled macro skills make coding agents cheap but brittle; primitive tool calls make them robust but costly. Here is how to run macros first, verify every result, and fall back to small steps only when the workspace shows it is needed.
Cutting LLM Inference Costs: Lessons from Cloudflare's 68% Cut
Cloudflare cut its AI code reviewer's average cost per review by 68%. Here is what that number does and does not mean, and a measured playbook for your own workload: prompt cache breakpoints, fewer agent calls, right-sized models and batch.
AI's Double-Edged Impact on DevOps: Faster Changes, Shakier Releases
DORA's research links AI adoption to lower delivery stability. A new simulation shows why: AI-written code overloads review and enlarges deployments, while CI agents can mask defects. Here is how to measure it and keep the speed.
Agentic AI in Production: Permissions, Approval Gates, Traces and Evals
A practical playbook for running AI agents in production: classify every tool by side effect, gate mutations behind a human, set hard stop conditions, trace each step with OpenTelemetry and evaluate continuously.
MCP for DevOps: Give Agents Pipeline Context Without the Keys
How platform teams can use the Model Context Protocol safely: start with read-only servers and scoped identities behind a gateway, trace every tool call, and add write tools only behind human approval. Includes RBAC, tool and task examples.
Gemini 4 Argon: What It Means for Your Team, and Where to Use It First
Gemini 4 Argon is built to work for hours on one measurable problem. Here is what shipped, the real costs, and how to pilot it safely on your own backlog.
Balinder WaliaLeer 
Workstation AI Shop: Promoting the AI Sales Storefront Through int → test → acc → prod
Workstation AI Shop sells GPU machines with a fully autonomous AI Sales Agent that configures the basket. See how Ring Promoter moves each build through int → test → acc → prod — with Orbit screenshots release engineers and stakeholders share.
Balinder WaliaLeer 
NVIDIA OpenShell: Secure Agent Runtime on Mac Studio and GPU Workstations
NVIDIA OpenShell governs autonomous agents with sandboxes, policy, and credential isolation. Workstation explains how to run it on Mac Studio (Apple GPU host inference) and on Linux with NVIDIA --gpu.
Balinder WaliaLeer 
Claude Code Routines: agentes en la nube programados, subrutinas y corrutinas
Las Routines de Claude Code ejecutan tareas de agente guardadas en la nube de Anthropic según una programación, mediante API o con eventos de GitHub. Cómo funcionan, cómo configurar una y cómo encadenarlas.
Balinder WaliaLeer 