agentleFS
Sign inStart free
milad-ahmdmilad-ahmd/agent-skills/skills/cost-cutter/SKILL.md

cost-cutter

Skill5 starsChanged 17 days agoMITView on GitHub

Finds ways to cut cloud and SaaS spend (AWS, GCP, Azure, Vercel, Netlify, Supabase, Firebase, OpenAI and other LLM APIs, databases, CDNs) by analyzing billing exports and infrastructure code, then ranks savings by dollars and effort with exact fixes and risks. Use when the user asks to reduce cloud costs, review a cloud bill, optimize infrastructure spend, cut LLM API costs, or asks why their bill is so high.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every recommendation comes with a dollar estimate, the effort and the risk.

    line 8
  2. Very specific:

    Use read-only CLI commands only if the user approves (for example aws ce get-cost-and-usage).

    line 15
  3. Distinctive:

    Never delete resources, data or backups without explicit confirmation and a backup plan.

    line 47
  4. Distinctive:

    Don't trade away reliability or security without stating the trade-off.

    line 48

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Cost Cutter 💸

Every recommendation comes with a dollar estimate, the effort and the risk.

1. Gather (read-only)

  • The bill: a cost export by service and usage type (AWS Cost Explorer CSV, GCP billing export, Vercel or Supabase usage pages, LLM usage dashboards). Ask for the last 1–3 months.
  • Infrastructure as code in the repo: Terraform, CDK, Pulumi, Kubernetes manifests, serverless.yml, vercel.json, Dockerfiles, CI workflows.
  • The code paths that drive cost: LLM calls (model, max tokens, retries, caching), queries, cron jobs, image and video processing, logging volume.
  • Use read-only CLI commands only if the user approves (for example aws ce get-cost-and-usage).

2. Checklist (common big wins)

AreaLook for
ComputeIdle/oversized instances, dev envs running 24/7, no autoscaling, x86 where ARM (Graviton) works, on-demand where Savings Plans/committed use fit steady load
StorageNo lifecycle rules, old snapshots/backups, unattached volumes, logs kept forever, S3 Standard for cold data
NetworkNAT gateway data processing, cross-AZ/region traffic, missing CDN caching, egress-heavy assets
DatabasesOversized instances, unused read replicas, missing indexes causing big scans, provisioned IOPS not needed
Serverless / edgeLong durations, over-provisioned memory, chatty functions, missing caching/ISR, bandwidth overages
LLM APIsLargest model for every task, no prompt caching, huge contexts, no max_tokens, retries without backoff, no batch API for offline jobs, missing response cache
ObservabilityLog/metric ingestion volume, debug logging in prod, high-cardinality metrics
SaaSUnused seats, overlapping tools, monthly vs annual plans

3. Report

## 💸 Savings report: est. $X/month (Y% of spend)
| # | Change | Est. saving/mo | Effort | Risk | How |
|---|---|---|---|---|---|
| 1 | Route classification calls to a smaller model | ~$420 | S | Low | change model in `src/ai/classify.ts` + eval 50 samples |

Order by saving ÷ effort. Show the estimation maths (usage × unit price) and label every number as an estimate. Prices change, so tell the user to confirm on the provider's pricing page.

4. Implement (only after approval)

One change at a time, through infrastructure-as-code or a PR, with a rollback note and a metric to watch (latency, error rate, quality evaluations for model changes).

Rules

  • Never delete resources, data or backups without explicit confirmation and a backup plan.
  • Don't trade away reliability or security without stating the trade-off.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free