Generates structured monthly cost reports for Claude API token usage with daily, project, and model breakdowns. Supports cross-month comparisons and cost optimization suggestions.
How this skill is triggered — by the user, by Claude, or both
Slash command
/claude-token-analyzer:cta-cost-auditThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Generate structured monthly cost reports with daily breakdown, project breakdown, and model cost comparison.
Generate structured monthly cost reports with daily breakdown, project breakdown, and model cost comparison.
Execute mcp__token-analyzer__sync_db. Skip if already called in this conversation.
Execute mcp__token-analyzer__cost_report with:
month: user-specified YYYY-MM, or current month by defaultdaily: trueproject_path: optional, only if user specifies a projectFormat results into the following structure. Fill every section from the cost_report response.
## CTA 月度成本報告 — YYYY-MM
**月度總成本:$X.XX USD**
### 每日成本
| 日期 | 成本 | 會話數 | 備註 |
|------|------|--------|------|
| 03-01 | $X.XX | N | |
| 03-05 | $X.XX | N | <- peak day |
### 按專案分解
| 專案 | 成本 | 佔比 |
|------|------|------|
| project-a | $X.XX | XX.X% |
### 按模型分解
| 模型 | 成本 | Token 數 | 每百萬 Token 均價 |
|------|------|----------|------------------|
| claude-opus-4-6 | $X.XX | X | $X.XX |
| claude-sonnet-4-6 | $X.XX | X | $X.XX |
| claude-haiku-4-5 | $X.XX | X | $X.XX |
### 優化建議
- (Calculate savings if Opus usage were replaced by Sonnet where applicable)
If the user asks about a specific project, execute mcp__token-analyzer__analyze_project with that project_path.
If the user requests comparison, call cost_report for the previous month and calculate:
$X.XX) and percentage (+X.X%).$X.XX USD.85.3%).125,000).${TMPDIR:-/tmp}/cta-cost-report.md, report path.For MCP tool parameter details: ${CLAUDE_PLUGIN_ROOT}/skills/cta/references/tool-reference.md
npx claudepluginhub li195111/claude-token-analyzer --plugin claude-token-analyzerGuides capturing AI tool costs from Anthropic and OpenAI dashboards quarterly, records structured Markdown data for health snapshots and MODEL_ROUTING.md updates.
Routes Claude Code token usage and cost analysis requests to sub-skills for health checks, cost audits, anomaly detection, project reviews, trends, and usage patterns. Provides quick overviews for ambiguous intents.