[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M74APKYFQXNTS5YXKDMCE0XB":3,"skill-same:01M74APKYFQXNTS5YXKDMCE0XB":60},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M74APKYFQXNTS5YXKDMCE0XB","claude-usage-report",null,"Aggregate Claude Code usage (tokens, sessions, API-equivalent cost) by day, week, month, model, and project from local JSONL session logs. Use when the user asks about their Claude Code usage, consumption, costs, \"cuánto he usado\", \"uso diario\u002Fsemanal\u002Fmensual\", or wants to audit which projects\u002Fmodels are eating the most tokens.","CarlosCaPe\u002Foctorato","CarlosCaPe","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F39838249?v=4","01M72WXX4ZRR8MBH1EP1QND2FQ","dataqbs","01M72WXX5NA7YCKRSJ54GW1GQ8",18,"MIT","https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fclaude-usage-report\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: claude-usage-report\ndescription: Aggregate Claude Code usage (tokens, sessions, API-equivalent cost) by day, week, month, model, and project from local JSONL session logs. Use when the user asks about their Claude Code usage, consumption, costs, \"cuánto he usado\", \"uso diario\u002Fsemanal\u002Fmensual\", or wants to audit which projects\u002Fmodels are eating the most tokens.\n---\n\n# Claude Usage Report\n\n## Purpose\n\nClaude Code stores every assistant message — including token usage and model\nidentity — in JSONL files under `~\u002F.claude\u002Fprojects\u002F`. This skill parses those\nlogs and produces a usage breakdown so the operator can:\n\n- See daily \u002F weekly \u002F monthly activity at a glance.\n- Identify which projects and models are driving cost.\n- Spot anomalies (sudden spike, idle weeks, model drift).\n- Quantify the \"value extracted\" from a Claude Max\u002FPro subscription vs paying\n  per-token at API list price.\n\n## Triggers\n\nActivate when the user asks any of:\n\n- \"¿cómo va mi uso de Claude?\" \u002F \"how is my Claude usage?\"\n- \"uso semanal\" \u002F \"uso mensual\" \u002F \"uso diario\"\n- \"cuánto he gastado en Claude\" \u002F \"Claude usage stats\"\n- \"qué proyecto consume más tokens\"\n- \"qué modelo uso más\"\n- Any audit-style question about Claude Code consumption.\n\n## Workflow\n\n### 1. Run the aggregator\n\n```bash\npython3 ~\u002F.claude\u002Fskills\u002Fclaude-usage-report\u002Fscripts\u002Fusage_report.py\n```\n\nOptional flags:\n\n- `--days 30` — show last 30 days of detail (default 14)\n- `--json` — emit machine-readable output (useful for piping into other tools)\n- `--projects-dir PATH` — override the default `~\u002F.claude\u002Fprojects\u002F` location\n\n### 2. Output sections (in order)\n\n1. **Header** — files scanned, lines, messages with usage data\n2. **ÚLTIMOS N DÍAS** — day-by-day breakdown (sessions, msgs, tokens, $)\n3. **POR SEMANA** — last 8 ISO weeks\n4. **POR MES** — every month with activity\n5. **POR MODELO** — sorted by cost, shows which model dominates spend\n6. **POR PROYECTO** — top 10 projects by cost (uses cwd-derived names)\n7. **TOTAL ACUMULADO** — grand totals + cost note\n\n### 3. Interpret with the user\n\nUse the output to answer their actual question. Common follow-ups worth\nvolunteering:\n\n- If Opus dominates >80% of cost: \"Sonnet handles ~80% of typical work at 1\u002F5\n  the price — worth considering for non-critical tasks.\"\n- If cache reads are large vs input tokens: \"Prompt caching is saving you ~X.\n  Keeping sessions warm matters.\"\n- If a single day is an extreme outlier: name it and check it's expected.\n- If sessions count is high but msgs\u002Fsession low: many short interruptions —\n  consolidating may help cache hit rate.\n\n## Pricing source\n\nThe script embeds Anthropic public list prices (USD per 1M tokens) for the\nClaude 4.x family. Pricing categories: `in`, `out`, `cache_w` (cache write),\n`cache_r` (cache read).\n\nWhen Anthropic updates prices or releases new models, edit the `PRICING` dict\nat the top of `scripts\u002Fusage_report.py`. Unknown models fall back to family\nheuristics (opus \u002F sonnet \u002F haiku) so the report still produces a number.\n\n## Best practices\n\n- **Don't conflate cost with payment.** With a Claude Max\u002FPro subscription the\n  dollar figure is \"what this would cost at API list price\" — i.e. value\n  extracted, not money owed. Always state this when reporting.\n- **Cache read is the cheapest token type** (~5% of input price) — high cache\n  read volume is good news, not bad.\n- **Synthetic messages** (model = `\u003Csynthetic>`) are harness-generated, not\n  billed. They appear in counts but with $0.\n- **Empty sessions** (msgs but no `usage` field) exist because Claude Code\n  sometimes writes user-only lines or aborted turns. The aggregator skips\n  records without `usage`, which is correct.\n\n## Error handling\n\n- **Empty `~\u002F.claude\u002Fprojects\u002F`** — script exits with a clear error. Suggest\n  running Claude Code at least once first.\n- **Malformed JSONL lines** — silently skipped per line (logs sometimes have\n  truncated tails after crashes).\n- **Unknown model name** — falls back to family heuristic via substring match.\n  If neither opus nor haiku is in the name, defaults to Sonnet pricing.\n\n## Lessons Learned\n\n- **Don't trust `npx` for one-off tools in restricted environments.** The\n  canonical community tool for this (`ccusage`) is great but is blocked by\n  Claude Code's auto-mode classifier (external code execution). Parsing the\n  JSONL directly with Python avoids that gate entirely.\n- **Project names come from `cwd`-derived directory names**, e.g.\n  `-home-user-Documents-github-foo`. The script strips the home prefix\n  generically (no hardcoded usernames) so the same code works for any operator.\n- **ISO weeks ≠ calendar weeks.** Use `datetime.isocalendar()` to match what\n  most billing dashboards report. Don't roll your own week math.\n- **`cache_creation_input_tokens` is expensive** (1.25× input price) — a\n  session that writes huge cache once then reads it many times is optimal;\n  one that re-writes cache every turn burns money.\n","# Claude Usage Report\n\n## Purpose\n\nClaude Code stores every assistant message — including token usage and model\nidentity — in JSONL files under `~\u002F.claude\u002Fprojects\u002F`. This skill parses those\nlogs and produces a usage breakdown so the operator can:\n\n- See daily \u002F weekly \u002F monthly activity at a glance.\n- Identify which projects and models are driving cost.\n- Spot anomalies (sudden spike, idle weeks, model drift).\n- Quantify the \"value extracted\" from a Claude Max\u002FPro subscription vs paying\n  per-token at API list price.\n\n## Triggers\n\nActivate when the user asks any of:\n\n- \"¿cómo va mi uso de Claude?\" \u002F \"how is my Claude usage?\"\n- \"uso semanal\" \u002F \"uso mensual\" \u002F \"uso diario\"\n- \"cuánto he gastado en Claude\" \u002F \"Claude usage stats\"\n- \"qué proyecto consume más tokens\"\n- \"qué modelo uso más\"\n- Any audit-style question about Claude Code consumption.\n\n## Workflow\n\n### 1. Run the aggregator\n\n```bash\npython3 ~\u002F.claude\u002Fskills\u002Fclaude-usage-report\u002Fscripts\u002Fusage_report.py\n```\n\nOptional flags:\n\n- `--days 30` — show last 30 days of detail (default 14)\n- `--json` — emit machine-readable output (useful for piping into other tools)\n- `--projects-dir PATH` — override the default `~\u002F.claude\u002Fprojects\u002F` location\n\n### 2. Output sections (in order)\n\n1. **Header** — files scanned, lines, messages with usage data\n2. **ÚLTIMOS N DÍAS** — day-by-day breakdown (sessions, msgs, tokens, $)\n3. **POR SEMANA** — last 8 ISO weeks\n4. **POR MES** — every month with activity\n5. **POR MODELO** — sorted by cost, shows which model dominates spend\n6. **POR PROYECTO** — top 10 projects by cost (uses cwd-derived names)\n7. **TOTAL ACUMULADO** — grand totals + cost note\n\n### 3. Interpret with the user\n\nUse the output to answer their actual question. Common follow-ups worth\nvolunteering:\n\n- If Opus dominates >80% of cost: \"Sonnet handles ~80% of typical work at 1\u002F5\n  the price — worth considering for non-critical tasks.\"\n- If cache reads are large vs input tokens: \"Prompt caching is saving you ~X.\n  Keeping sessions warm matters.\"\n- If a single day is an extreme outlier: name it and check it's expected.\n- If sessions count is high but msgs\u002Fsession low: many short interruptions —\n  consolidating may help cache hit rate.\n\n## Pricing source\n\nThe script embeds Anthropic public list prices (USD per 1M tokens) for the\nClaude 4.x family. Pricing categories: `in`, `out`, `cache_w` (cache write),\n`cache_r` (cache read).\n\nWhen Anthropic updates prices or releases new models, edit the `PRICING` dict\nat the top of `scripts\u002Fusage_report.py`. Unknown models fall back to family\nheuristics (opus \u002F sonnet \u002F haiku) so the report still produces a number.\n\n## Best practices\n\n- **Don't conflate cost with payment.** With a Claude Max\u002FPro subscription the\n  dollar figure is \"what this would cost at API list price\" — i.e. value\n  extracted, not money owed. Always state this when reporting.\n- **Cache read is the cheapest token type** (~5% of input price) — high cache\n  read volume is good news, not bad.\n- **Synthetic messages** (model = `\u003Csynthetic>`) are harness-generated, not\n  billed. They appear in counts but with $0.\n- **Empty sessions** (msgs but no `usage` field) exist because Claude Code\n  sometimes writes user-only lines or aborted turns. The aggregator skips\n  records without `usage`, which is correct.\n\n## Error handling\n\n- **Empty `~\u002F.claude\u002Fprojects\u002F`** — script exits with a clear error. Suggest\n  running Claude Code at least once first.\n- **Malformed JSONL lines** — silently skipped per line (logs sometimes have\n  truncated tails after crashes).\n- **Unknown model name** — falls back to family heuristic via substring match.\n  If neither opus nor haiku is in the name, defaults to Sonnet pricing.\n\n## Lessons Learned\n\n- **Don't trust `npx` for one-off tools in restricted environments.** The\n  canonical community tool for this (`ccusage`) is great but is blocked by\n  Claude Code's auto-mode classifier (external code execution). Parsing the\n  JSONL directly with Python avoids that gate entirely.\n- **Project names come from `cwd`-derived directory names**, e.g.\n  `-home-user-Documents-github-foo`. The script strips the home prefix\n  generically (no hardcoded usernames) so the same code works for any operator.\n- **ISO weeks ≠ calendar weeks.** Use `datetime.isocalendar()` to match what\n  most billing dashboards report. Don't roll your own week math.\n- **`cache_creation_input_tokens` is expensive** (1.25× input price) — a\n  session that writes huge cache once then reads it many times is optimal;\n  one that re-writes cache every turn burns money.\n",{"name":5,"description":7},"master","b4414608cd5a3f87d1525b518f1c6add8363540d","skills\u002Fclaude-usage-report",6,"https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato","https:\u002F\u002Fwww.dataqbs.com\u002Foctorato","https:\u002F\u002Fgithub.com\u002FCarlosCaPe","www.dataqbs.com"," AI-powered data engineering\r\n& consulting\r\n\r\nWe build data platforms that turn complexity into clarity — from cloud architecture and ETL pipelines to AI-driven","2026-11-14 00:21:00",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add CarlosCaPe\u002Foctorato --skill claude-usage-report",{"tool":40,"command":41},"bunx","bunx skills add CarlosCaPe\u002Foctorato --skill claude-usage-report",{"tool":43,"command":44},"pnpm","pnpm dlx skills add CarlosCaPe\u002Foctorato --skill claude-usage-report","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fclaude-usage-report\u002FSKILL.md",true,[48,54],{"path":49,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":53},"scripts\u002Fusage_report.py","SCRIPT","脚本",10590,"https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fclaude-usage-report\u002Fscripts\u002Fusage_report.py",{"path":55,"resourceType":56,"resourceTypeDesc":57,"sizeBytes":58,"sourceHtmlUrl":59},"skill.json","REFERENCE","参考文档",452,"https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fclaude-usage-report\u002Fskill.json",[61,70,78,85,93,101],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":65,"licenseFileUrl":66,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":19},"01M74EVAA4A4B4AKR5TT732GP4","yeet","Use only when the user explicitly asks to stage, commit, push, and open a GitHub pull request in one flow using the GitHub CLI (`gh`).","Apache-2.0","https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fyeet\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fyeet\u002FSKILL.md",[],[],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":65,"licenseFileUrl":74,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":19},"01M74EV4WMRDKP2NQS4YWR8PXH","wrangler","Cloudflare Workers CLI for deploying, developing, and managing Workers, KV, R2, D1, Vectorize, Hyperdrive, Workers AI, Containers, Queues, Workflows, Pipelines, and Secrets Store. Load before running wrangler commands to ensure correct syntax and best practices. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.","https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fwrangler\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fwrangler\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":19},"01M74EV05MHTXKEXD4B2QJEYTF","workspace-skill-discovery","Discover and include ALL skills across the workspace — not just global ones. Use at the start of every session or when the user asks about available skills. Ensures project-level skills (under .claude\u002Fskills\u002F or skills\u002F in any repo) are never missed.","https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fworkspace-skill-discovery\u002FSKILL.md",[],[],{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":65,"licenseFileUrl":89,"sourceHtmlUrl":90,"categories":91,"keywords":92,"publishedAt":19},"01M74ETV78525B4VP2F28Z6NRD","workers-best-practices","Reviews and authors Cloudflare Workers code against production best practices. Load when writing new Workers, reviewing Worker code, configuring wrangler.jsonc, or checking for common Workers anti-patterns (streaming, floating promises, global state, secrets, bindings, observability). Biases towards retrieval from Cloudflare docs over pre-trained knowledge.","https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fworkers-best-practices\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fworkers-best-practices\u002FSKILL.md",[],[],{"skillNo":94,"name":95,"nameZh":6,"description":96,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":65,"licenseFileUrl":97,"sourceHtmlUrl":98,"categories":99,"keywords":100,"publishedAt":19},"01M74ETR75NF67WGVEN7288ACV","winui-app","Create, build and design WinUI 3 desktop apps with C# and the Windows App SDK following Microsoft's official guidance and CommunityToolkit. Covers XAML, navigation, theming, accessibility and deployment.","https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fwinui-app\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fwinui-app\u002FSKILL.md",[],[],{"skillNo":102,"name":103,"nameZh":6,"description":104,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":105,"categories":106,"keywords":107,"publishedAt":19},"01M74ETJFRNPPNSQZ36C6HD6CW","windows-brain-script-portability","Scripts in ~\u002F.claude\u002Fscripts\u002F that run on a Windows brain: use the python3.cmd shim and not bare python3, and reconfigure stdout to UTF-8 if you print symbols. Use it when writing a brain script or triaging a Windows-only crash.","https:\u002F\u002Fgithub.com\u002FCarlosCaPe\u002Foctorato\u002Fblob\u002Fmaster\u002Fskills\u002Fwindows-brain-script-portability\u002FSKILL.md",[],[]]