cost-review — Claude Code skill: cost analysis
cost-review — a Claude Code skill that breaks down platform compute and storage costs for a chosen period. It builds a monthly review by job, pipeline, and table.
MIT ★ 106 updated 2026-11-06
/plugin marketplace add DataWithBaraa/claude-code-data-engineering-skills
/plugin install data-engineering@data-with-baraaWhat cost-review does
The skill turns a period’s spending into one HTML page. It reads billing and usage records, run records, query logs, and the catalog. The default period is the previous month.
The report lists the costliest jobs and their outputs, scheduled pipeline cost per run, storage charges, and idle resources. It flags spending spikes with the day, job, and change behind each one. Proposals are ranked by monthly savings and risk. The list of unread tables comes from the big-review skill.
It proposes changes but takes no action: it will not delete tables, stop jobs, or modify clusters. The output is a single HTML file.
What it helps with
- identify jobs with the highest costs
- compare scheduled pipeline run costs across periods
- find spending on idle clusters and warehouses
- trace a cost spike to its day, job, and triggering change
- rank cost-cutting proposals by savings and risk
Example prompts for the agent
- Review the platform’s spending for last month.
- Break down costs for
2025-03by job, pipeline, and table. - Find the cost spikes and suggest savings for this period.
How to install cost-review
1Plugin marketplace
/plugin marketplace add DataWithBaraa/claude-code-data-engineering-skills/plugin install data-engineering@data-with-baraaThis installs the whole data-engineering plugin with all its skills
2Skills CLI (Vercel)
npx skills add DataWithBaraa/claude-code-data-engineering-skills --skill cost-review -a claude-code3Manually: put the files in place
~/.claude/skills/cost-review/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/cost-review/
1Skills CLI (Vercel)
npx skills add DataWithBaraa/claude-code-data-engineering-skills --skill cost-review -a codex2Manually: put the files in place
.agents/skills/cost-review/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add DataWithBaraa/claude-code-data-engineering-skills --skill cost-review -a cursor2Manually: put the files in place
.cursor/skills/cost-review/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install DataWithBaraa/claude-code-data-engineering-skills skills/cost-review2Manually: put the files in place
.github/skills/cost-review/SKILL.md1Skills CLI (Vercel)
npx skills add DataWithBaraa/claude-code-data-engineering-skills --skill cost-review -a gemini-cli2Platform CLI
gemini skills install https://github.com/DataWithBaraa/claude-code-data-engineering-skills --path skills/cost-reviewManually: put the files in place
.devin/skills/cost-review/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about cost-review
How do I install cost-review in Claude Code?
Add the marketplace with /plugin marketplace add DataWithBaraa/claude-code-data-engineering-skills, then install the skill with /plugin install data-engineering@data-with-baraa.
What does cost-review do?
It creates an HTML breakdown of compute and storage costs for a period. The report covers costly jobs, pipeline runs, idle resources, and spending spikes, then lists proposals ranked by savings and risk.
Where does the unread-table list come from?
The list is supplied by the big-review skill.
Source text
SKILL.md · MIT
--- name: cost-review description: Use for the monthly cost review of the platform, to see where the compute and the storage money goes, per job, pipeline and table, and what to cut. argument-hint: "[period]" --- # Cost review The job: where the money goes, per job, per pipeline and per table, and what to do about it. It proposes, it stops and deletes nothing. Input: the period, the last month by default. ## Steps 1. Read the billing and usage records of the period: compute per job and pipeline, storage per table. 2. The top 10 jobs by cost, and what each one produces. 3. The cost per run of each scheduled pipeline, against the previous period. 4. The storage bill: what the storage costs per month, and the part spent on tables nobody reads. The list of those tables comes from the big-review skill. 5. Idle spend: clusters and warehouses that ran with nothing to do, and jobs that ran for nobody. 6. The spikes: which day, which job, and which change caused it. 7. Proposals, each with the saving per month and the risk, in order of the saving. 8. Give the result as one web page. ## Output One HTML file: the cost of the period on top, the top jobs, the cost per run, the storage bill, the idle spend, the spikes, and the proposals last. ## Reads - the billing and usage records, the run records, the query logs, the catalog ## Writes - the web page only ## Never - Never deletes a table, never stops a job, never changes a cluster: it proposes.
© Data With Baraa · MIT · DataWithBaraa/claude-code-data-engineering-skills