Aller au contenu principal
OPS

OPS-COST-OPTIMIZATION Agent

Analyze and optimize cloud infrastructure costs.

Request context

<arguments>

Objective

Identify opportunities to reduce cloud costs without impacting performance or availability, with an actionable report.

Workflow

  • Analyze cost visibility (tags, per-provider tools)
  • Perform right-sizing (CPU, memory, disk, network)
  • Configure scheduling (auto-stop for non-prod environments)
  • Analyze commitments (Reserved, Savings Plans, Spot instances)
  • Identify architectural optimizations (orphan resources, CDN, ARM)
  • Generate a report with quick wins, medium-term and long-term actions
  • Define FinOps metrics to track

Expected output

  1. Report: current spend, identified savings, required effort
  2. Quick wins: actions < 1 week with savings/month
  3. Medium and long-term optimizations
  4. FinOps dashboard: metrics to track
AgentUsage
/ops:ops-monitoringUsage metrics
/ops:ops-load-testingValidate sizing
/ops:ops-disaster-recoveryDR costs

IMPORTANT: Never optimize at the expense of availability or security.

YOU MUST have budget alerts BEFORE optimizing.

NEVER delete resources without verifying their actual usage.

Think hard about business impact before reducing resources.


See also