Requirements
- Target platform
- OpenClaw
- Install method
- Manual import
- Extraction
- Extract archive
- Prerequisites
- OpenClaw
- Primary doc
- SKILL.md
Maximize your Claude Code subscription value with smart usage monitoring and burn rate optimization. Track 5-hour session and 7-day weekly quotas, get one-ti...
Maximize your Claude Code subscription value with smart usage monitoring and burn rate optimization. Track 5-hour session and 7-day weekly quotas, get one-ti...
Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.
I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Then review README.md for any prerequisites, environment setup, or post-install checks. Tell me what you changed and call out any manual steps you could not complete.
I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run.
Version: 1.0.5 Get the most out of your Claude Code subscription by monitoring usage in real-time and optimizing your daily burn rate.
You're paying $20-200/month for Claude Code. Are you: β Using it to its full potential? β Hitting limits too early? β Leaving quota unused at reset? This skill tracks your 5-hour session and 7-day weekly quotas, calculates your daily burn rate, and tells you if you should use more or throttle back.
π Burn Rate Tracking β Are you under/over/on-pace for optimal usage? β‘ Smart Alerts β One-time warnings when SESSION > 50% (no spam) π― Plan-Aware β Auto-detects Pro ($20), Max 100 ($100), Max 200 ($200) πΎ Ultra-Lightweight β 10-minute cache, minimal API calls π Daily Reports β Evening summary: SESSION, WEEKLY, burn rate π Token Health Check β Hourly check + alert if manual refresh needed (~once per week)
Run the setup wizard to configure your OAuth tokens: cd {baseDir} ./scripts/setup.sh You'll need: Access Token (sk-ant-oat01-...) Refresh Token (sk-ant-ort01-...) See references/token-extraction.md for how to get these.
./scripts/check-usage.sh Output: SESSION=22.0 WEEKLY=49.0 BURN_RATE=OK CACHED_AT=1771583780
./scripts/report.sh Output: π Claude Code Daily Check: β±οΈ SESSION (5h): 22% π WEEKLY (7d): 49% βͺ ΠΠ° ΡΠ΅ΠΌΠΏΡ β ΠΎΠΏΡΠΈΠΌΠ°Π»ΡΠ½Π΅ Π²ΠΈΠΊΠΎΡΠΈΡΡΠ°Π½Π½Ρ
π’ UNDER β You're under-using your subscription. Use more to get your money's worth! βͺ OK β On pace. Optimal usage for your plan. π΄ OVER β Over-burning. You'll hit limits before reset.
PlanMonthlyWeekly BudgetDaily BudgetPro$20~14%~2%Max 100$100~14%~2%Max 200$200~14%~2% (7-day window resets weekly, so ~14% per day = 100% per week)
SESSION > 50% β π‘ One-time warning (won't repeat until next reset) WEEKLY > 80% β π‘ One-time warning Alerts use state tracking (/tmp/claude-usage-alert-state) to avoid spam.
Default: /tmp/claude-usage.cache with 10-minute TTL. Override: CACHE_FILE=/custom/path CACHE_TTL=300 ./scripts/check-usage.sh
scripts/setup.sh β Initial token configuration scripts/check-usage.sh β Core usage checker (cached, burn rate calc) scripts/report.sh β Human-readable daily report references/api-endpoint.md β Anthropic OAuth API docs references/token-extraction.md β How to get OAuth tokens references/plans.md β Claude Code subscription tiers
GET https://api.anthropic.com/api/oauth/usage Authorization: Bearer <access-token> anthropic-beta: oauth-2025-04-20 Response: { "five_hour": { "utilization": 22.0, "resets_at": "2026-02-20T14:00:00.364238+00:00" }, "seven_day": { "utilization": 49.0, "resets_at": "2026-02-24T10:00:01.364256+00:00" } }
curl β API requests date β Timestamp parsing grep, cut, printf β Text parsing No external dependencies (jq, etc.).
Tokens are stored in {baseDir}/.tokens (gitignored). Never share your access/refresh tokens.
OAuth tokens work for ~1 week, then need manual refresh. Set up 30-minute health check for better reliability: # Add cron job to check token health every 30 minutes openclaw cron add \ --name "claude-token-refresh" \ --every 30m \ --announce \ --message "ΠΠ°ΠΏΡΡΡΠΈ {baseDir}/scripts/auto-refresh-cron.sh" What it does: β Token valid β silent (no spam) π΄ Token expired β one-time alert with manual refresh instructions Manual refresh (once per week, 30 seconds): claude auth login # Browser opens β sign in to claude.ai β done! Tokens auto-sync to {baseDir}/.tokens after successful login.
"No token configured" β Run ./scripts/setup.sh "Token expired" / "API request failed" β OAuth tokens expire after ~1 week β Manual refresh: claude auth login (browser opens β sign in β done) β Set up hourly health check to get alerts before expiry (see above) Burn rate shows empty β API response missing resets_at β try again in a few minutes Auto-refresh failed β OAuth refresh endpoint may have changed β Manual refresh: claude auth login β copy new tokens β run ./scripts/setup.sh
π Bugfix: Fixed token extraction in auto-refresh-cron.sh (removed quotes handling) β‘ Performance: Reduced cron interval from 1h to 30m for more reliable token refresh π Improved reliability of OAuth token sync with ~/.claude/.credentials.json
π Replaced automatic refresh with health check + manual refresh workflow π Updated documentation with manual refresh instructions β° Health check alerts when manual refresh needed (~once per week)
β±οΈ Fixed auto-refresh interval (hourly instead of 5h) π Improved burn rate calculation accuracy
Found a bug or have a feature request? β Open an issue on ClawHub: https://clawhub.ai/friday/token-usage-optimizer
MIT
Workflow acceleration for inboxes, docs, calendars, planning, and execution loops.
Largest current source with strong distribution and engagement signals.