# Send Clawmeter to your agent Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually. ## Fast path - Download the package from Yavira. - Extract it into a folder your agent can access. - Paste one of the prompts below and point your agent at the extracted folder. ## Suggested prompts ### New install ```text 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. ``` ### Upgrade existing ```text 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. ``` ## Machine-readable fields ```json { "schemaVersion": "1.0", "item": { "slug": "clawmeter", "name": "Clawmeter", "source": "tencent", "type": "skill", "category": "开发工具", "sourceUrl": "https://clawhub.ai/Someone0070/clawmeter", "canonicalUrl": "https://clawhub.ai/Someone0070/clawmeter", "targetPlatform": "OpenClaw" }, "install": { "downloadUrl": "/downloads/clawmeter", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=clawmeter", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "packageFormat": "ZIP package", "primaryDoc": "SKILL.md", "includedAssets": [ "CHANGELOG.md", "CONTRIBUTING.md", "PACKAGE_SUMMARY.md", "PUBLICATION_READY.md", "PUBLISHING.md", "README.md" ], "downloadMode": "redirect", "sourceHealth": { "source": "tencent", "slug": "clawmeter", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-05-01T19:07:42.497Z", "expiresAt": "2026-05-08T19:07:42.497Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=clawmeter", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=clawmeter", "contentDisposition": "attachment; filename=\"clawmeter-1.0.0.zip\"", "redirectLocation": null, "bodySnippet": null, "slug": "clawmeter" }, "scope": "item", "summary": "Item download looks usable.", "detail": "Yavira can redirect you to the upstream package for this item.", "primaryActionLabel": "Download for OpenClaw", "primaryActionHref": "/downloads/clawmeter" }, "validation": { "installChecklist": [ "Use the Yavira download entry.", "Review SKILL.md after the package is downloaded.", "Confirm the extracted package contains the expected setup assets." ], "postInstallChecks": [ "Confirm the extracted package includes the expected docs or setup files.", "Validate the skill or prompts are available in your target agent workspace.", "Capture any manual follow-up steps the agent could not complete." ] } }, "links": { "detailUrl": "https://openagent3.xyz/skills/clawmeter", "downloadUrl": "https://openagent3.xyz/downloads/clawmeter", "agentUrl": "https://openagent3.xyz/skills/clawmeter/agent", "manifestUrl": "https://openagent3.xyz/skills/clawmeter/agent.json", "briefUrl": "https://openagent3.xyz/skills/clawmeter/agent.md" } } ``` ## Documentation ### ClawMeter — Cost Tracking Dashboard Track OpenClaw API usage and spending in real-time with budget alerts. ### Overview ClawMeter is a self-hosted cost tracking dashboard that monitors your OpenClaw session logs, calculates token usage and API costs, and provides real-time analytics with customizable budget alerts. Perfect for: Users who want visibility into their AI spending Teams managing OpenClaw deployments with budgets Developers optimizing model selection based on cost Anyone who wants to avoid surprise API bills ### Prerequisites OpenClaw installed and running Node.js v18+ (for the dashboard server) Session logs enabled in OpenClaw (default) ### Setup Clone or download ClawMeter: cd ~/.openclaw/workspace git clone https://github.com/yourusername/clawmeter.git cd clawmeter Install dependencies: npm install Configure (optional): cp .env.example .env # Edit .env to customize paths, budget limits, or alert settings Ingest existing logs: npm run ingest Start the dashboard: npm start Open dashboard: Navigate to http://localhost:3377 ### Commands When this skill is installed, your OpenClaw agent can respond to: ### Spending Queries "How much have I spent today?" Shows today's total spend and budget percentage "What's my monthly spend?" Displays current month total and budget status "Show me this week's costs" Summarizes last 7 days of spending ### Cost Analysis "Which model is most expensive?" Breaks down costs by model (Claude, GPT, Gemini, etc.) "Show my most expensive sessions" Lists top sessions by total cost "What's my average daily spend?" Calculates mean daily cost over the last 30 days ### Dashboard & Reports "Open my cost dashboard" Provides link to http://localhost:3377 "Generate a cost report" Exports summary data (today/week/month/all-time) ### Data Management "Refresh cost data" Triggers manual re-ingest of session logs "Clear old cost data" (Optional) Archive or delete data older than X days ### API Endpoints ClawMeter exposes a REST API that agents can query: ### Summary Statistics GET http://localhost:3377/api/summary Response: { "today": 2.15, "week": 8.42, "month": 32.76, "allTime": 127.89, "sessions": 234, "messages": 1842, "budgetDaily": 5.0, "budgetMonthly": 100.0 } ### Daily Breakdown GET http://localhost:3377/api/daily?days=30 Response: [ { "date": "2026-02-14", "cost": 2.15, "input_tokens": 45820, "output_tokens": 12340, "messages": 18 } ] ### Sessions GET http://localhost:3377/api/sessions?limit=50&offset=0 Response: [ { "id": "abc123", "agent": "main", "model": "claude-sonnet-4-5", "total_cost": 0.842, "message_count": 12 } ] ### Model Breakdown GET http://localhost:3377/api/models Response: [ { "model": "claude-sonnet-4-5", "provider": "anthropic", "total_cost": 45.62, "message_count": 324 } ] ### Top Sessions GET http://localhost:3377/api/top-sessions?limit=10 ### Budget Alerts GET http://localhost:3377/api/alerts ### Manual Ingest POST http://localhost:3377/api/ingest ### Environment Variables Edit .env to customize: # Paths OPENCLAW_AGENTS_DIR=/home/youruser/.openclaw/agents CLAWMETER_DB=/home/youruser/.openclaw/workspace/clawmeter/data/clawmeter.db # Server PORT=3377 # Budgets (USD) BUDGET_DAILY_LIMIT=5.00 BUDGET_MONTHLY_LIMIT=100.00 # Telegram alerts (optional) TELEGRAM_BOT_TOKEN=your_bot_token TELEGRAM_CHAT_ID=your_chat_id # Email alerts (optional) SMTP_HOST=smtp.gmail.com SMTP_PORT=587 SMTP_USER=your@email.com SMTP_PASS=your_app_password ALERT_EMAIL_TO=notify@yourdomain.com ### Budget Alerts ClawMeter automatically checks spending against configured thresholds and sends alerts via: Telegram — Instant notifications to your phone Email — Delivered via SMTP Alert triggers: Daily spend exceeds BUDGET_DAILY_LIMIT Monthly spend exceeds BUDGET_MONTHLY_LIMIT Alert throttling: One alert per type per day (prevents spam). ### Example 1: Daily Spend Check User: "How much have I spent today?" Agent Response: Let me check your spending... 📊 Today's Spend: $2.15 Budget: $5.00 (43% used) You have $2.85 remaining today. Implementation: const res = await fetch('http://localhost:3377/api/summary'); const data = await res.json(); const pct = (data.today / data.budgetDaily * 100).toFixed(0); return \`📊 Today's Spend: $${data.today.toFixed(2)}\\nBudget: $${data.budgetDaily.toFixed(2)} (${pct}% used)\\n\\nYou have $${(data.budgetDaily - data.today).toFixed(2)} remaining today.\`; ### Example 2: Model Comparison User: "Which model should I use to save money?" Agent Response: Based on your usage history: 💰 Cost per 1K messages: • Claude Haiku 3-5: $0.12 • GPT-4o-mini: $0.18 • Claude Sonnet 4-5: $0.45 • Claude Opus 4: $1.23 Recommendation: Use Haiku for simple tasks, Sonnet for complex work. ### Example 3: Budget Warning Agent (proactive): ⚠️ Budget Alert You've spent $5.52 today, exceeding your $5.00 daily limit. Top sessions today: 1. Code review task — $2.10 2. Research project — $1.85 3. Email drafting — $0.98 Consider switching to a cheaper model or pausing non-urgent tasks. ### Overview Screen Today's spend with budget progress bar Weekly and monthly totals All-time spending and message count Budget status (green/yellow/red indicators) ### Analytics Daily cost chart (bar chart, 7/14/30/90 day views) Model breakdown (donut chart with percentages) Top sessions by cost Recent activity feed ### Alerts History View all triggered budget alerts See when and why alerts fired Track spending patterns over time ### 1. Personal Budget Management Scenario: You want to keep OpenClaw costs under $100/month. Setup: Set BUDGET_MONTHLY_LIMIT=100.00 Enable Telegram alerts Check dashboard weekly Result: Get notified before you exceed budget, adjust usage accordingly. ### 2. Team Cost Allocation Scenario: Multiple team members use shared OpenClaw instance. Setup: ClawMeter tracks costs by agent/session Export data via API for reporting Analyze which agents/projects cost most Result: Fair cost distribution, identify optimization opportunities. ### 3. Model Selection Optimization Scenario: You're unsure which model to use for different tasks. Setup: Run experiments with different models Compare costs in ClawMeter dashboard Analyze cost-per-message and cost-per-token Result: Data-driven model selection, optimize for cost/quality balance. ### 4. Production Monitoring Scenario: OpenClaw running in production, need cost visibility. Setup: ClawMeter auto-ingests logs in real-time Set aggressive budget alerts Monitor dashboard for anomalies Result: Catch cost spikes immediately, prevent runaway spending. ### Dashboard shows $0.00 Cause: Session logs not found or not yet ingested. Fix: Verify OPENCLAW_AGENTS_DIR points to correct path Run npm run ingest manually Check that .jsonl files exist in agents/*/sessions/ ### Auto-watch not working Cause: File watcher not detecting changes. Fix: Ensure server is running (npm start) Check terminal for 📥 X new events messages Verify file permissions on session directories ### Budget alerts not sending Cause: Missing or incorrect credentials. Fix: Double-check TELEGRAM_BOT_TOKEN and TELEGRAM_CHAT_ID in .env For email, verify SMTP settings Check server logs for error messages ### Database errors Cause: Corrupted or locked database file. Fix: Stop the server Backup data/clawmeter.db Delete database and re-run npm run ingest ### Custom Model Pricing If you use custom models or pricing changes: Edit src/pricing.mjs Add entry to MODEL_PRICING object: 'your-custom-model': { input: 2.50, // per million tokens output: 10.00, // per million tokens cacheRead: 0.25, // optional cacheWrite: 3.75 // optional } Restart server ### Scheduled Reports Use cron to schedule daily/weekly reports: # Daily summary at 9 AM 0 9 * * * curl -s http://localhost:3377/api/summary | mail -s "Daily Cost Report" you@example.com ### Data Export Export raw data for external analysis: # Export all sessions to CSV sqlite3 data/clawmeter.db "SELECT * FROM sessions" -csv > sessions.csv ### Running as a Service Create systemd service for persistent operation: # /etc/systemd/system/clawmeter.service [Unit] Description=ClawMeter Cost Tracking Dashboard After=network.target [Service] Type=simple User=youruser WorkingDirectory=/home/youruser/.openclaw/workspace/clawmeter ExecStart=/usr/bin/npm start Restart=always [Install] WantedBy=multi-user.target Enable and start: sudo systemctl enable clawmeter sudo systemctl start clawmeter ### Local Access Only By default, ClawMeter binds to localhost:3377 and is not exposed to the internet. To access remotely, use SSH tunneling: ssh -L 3377:localhost:3377 your-server ### Sensitive Data Session logs may contain conversation data. ClawMeter: Stores only usage metadata (tokens, costs, timestamps) Does not store message content or prompts Database is local SQLite (no cloud sync) ### API Authentication ClawMeter currently has no authentication. For production: Run behind a reverse proxy (nginx/caddy) with auth Use firewall rules to restrict access Deploy on a private network ### Resource Usage CPU: Minimal (event-driven ingestion) RAM: ~50-100 MB Disk: Database grows ~1 KB per message (SQLite is very efficient) Network: None (all local) ### Scalability ClawMeter handles: 100K+ messages easily Years of historical data without performance degradation Multiple agents (separated by directory structure) For massive deployments (millions of messages), consider: Periodic database archiving PostgreSQL backend (requires code changes) ### Roadmap Planned features: Multi-user authentication PostgreSQL support for large deployments Export to CSV/JSON Cost forecasting (predict monthly spend) Slack/Discord webhook support Token usage heatmaps Model performance tracking (latency, quality) Budget recommendations based on usage patterns ### Support Documentation: GitHub README Issues: GitHub Issues Community: OpenClaw Discord ### License MIT — Free and open source forever. Made with ⚡ by the OpenClaw community Track your spending. Optimize your costs. Stay in control. ## Trust - Source: tencent - Verification: Indexed source record - Publisher: Someone0070 - Version: 1.0.0 ## Source health - Status: healthy - Item download looks usable. - Yavira can redirect you to the upstream package for this item. - Health scope: item - Reason: direct_download_ok - Checked at: 2026-05-01T19:07:42.497Z - Expires at: 2026-05-08T19:07:42.497Z - Recommended action: Download for OpenClaw ## Links - [Detail page](https://openagent3.xyz/skills/clawmeter) - [Send to Agent page](https://openagent3.xyz/skills/clawmeter/agent) - [JSON manifest](https://openagent3.xyz/skills/clawmeter/agent.json) - [Markdown brief](https://openagent3.xyz/skills/clawmeter/agent.md) - [Download page](https://openagent3.xyz/downloads/clawmeter)