AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

cost-optimization

Optimize existing Azure resources and analyze Reservations or Savings Plans. WHEN: "optimize Azure costs", "reduce cloud spending", "rightsize resources", "find idle resources", "orphaned disk", "deleted VM still charged", "public IP still charging", "reservation utilization", "Savings Plan coverage", "commitment recommendation", "why is pay-as-you-go still charged". DO NOT USE FOR: cost spikes, forecasts, pricing estimates, budgets, or governance.

Original source
microsoft
License
MIT
Pinned revision
3f5b461b2314e01c400564051e92469114c5b68f

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

Ask your assistant

Use Agentic Memory skills to read microsoft--cost-optimization at revision 3f5b461b2314e01c400564051e92469114c5b68f, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 9

references/commitments.md

Commitment Analysis

Resolve Scope and Terms

  1. Require an accessible billing scope for portfolio utilization. If the user supplies only a subscription, resolve or ask for its billing account and billing profile before calling list_benefit_utilization.
  2. Keep Reservations and Savings Plans separate. Report commitment name, term, scope, region or service constraints, quantity or hourly commitment, utilization period, currency, and expiration when available.
  3. Distinguish:
    • utilization: how much purchased commitment was consumed;
    • coverage: how much eligible usage received commitment pricing;
    • savings: tool-reported benefit for its stated period. High utilization can coexist with low coverage, and low utilization does not prove that a new purchase is needed.
  4. When the target usage includes AKS, check the host skill inventory and invoke aks-cost-optimization for its commitment-demand handoff when available. Reuse its time-aligned SKU or compatible compute group, region and zone, Spot/on-demand capacity type, observed hours and node counts, lookback coverage, exact resource mapping, planned changes, CRG impact, variability, and evidence gaps. If representative compute history is missing, report stable demand as unverified and do not validate an AKS commitment recommendation. Do not install the optional plugin without user consent.

Analyze Current Commitments

  1. Report utilization per commitment. Treat less than 90% as an investigation signal, not an automatic exchange or refund recommendation.
  2. Call query_costs with AmortizedCost, grouped by PricingModel, for the same period. Report committed and eligible on-demand spend per currency. Compare actual and amortized views when a recent purchase distorts cash cost.
  3. For low utilization, inspect list_reservation_transactions over the supported lookback and correlate purchases, refunds, or exchanges with scope, region, family, and resource-inventory changes. Missing history is inconclusive.

Evaluate New Recommendations

  1. Call get_benefit_recommendations. Preserve lookback, term, scope, savings amount and percentage, coverage, and expected utilization. Do not relabel period savings as monthly savings or combine currencies.
  2. Check whether rightsizing, shutdowns, migrations, expiring commitments, or policy restrictions would invalidate the recommendation. Present exchange, scope adjustment, workload alignment, expiration planning, or a new purchase as distinct options with evidence and tradeoffs.

Report and Safety

Present utilization and coverage separately, eligible on-demand spend, current commitment risks, new recommendations, confidence, and the next review date. Never purchase, exchange, or refund without explicit approval.

If a benefit or reservation tool is unavailable, use the API mapped in tool fallback. Preserve its period and scope semantics.