claude-usage-reportlisted

# Claude Usage Report ## Purpose Claude Code stores every assistant message — including token usage and model identity — in JSONL files under `~/.claude/projects/`. This skill parses those logs and produces a usage breakdown so the operator can: - See daily / weekly / monthly activity at a glance. - Identify which projects and models are driving cost. - Spot anomalies (sudden spike, idle weeks, model drift). - Quantify the "value extracted" from a Claude Max/Pro subscription vs paying per-token at API list price. ## Triggers Activate when the user asks any of: - "¿cómo va mi uso de Claude?" / "how is my Claude usage?" - "uso semanal" / "uso mensual" / "uso diario" - "cuánto he gastado en Claude" / "Claude usage stats" - "qué proyecto consume más tokens" - "qué modelo uso más" - Any audit-style question about Claude Code consumption. ## Workflow ### 1. Run the aggregator ```bash python3 ~/.claude/skills/claude-usage-report/scripts/usage_report.py ``` Optional flags: - `--days 30` — show last 30 days of detail (default 14) - `--json` — emit machine-readable output (useful for piping into other tools) - `--projects-dir PATH` — override the default `~/.claude/projects/` location ### 2. Output sections (in order) 1. **Header** — files scanned, lines, messages with usage data 2. **ÚLTIMOS N DÍAS** — day-by-day breakdown (sessions, msgs, tokens, $) 3. **POR SEMANA** — last 8 ISO weeks 4. **POR MES** — every month with activity 5. **POR MODELO** — sorted by cost, shows