{ "schemaVersion": "1.0", "item": { "slug": "token-watch", "name": "Token Watch", "source": "tencent", "type": "skill", "category": "数据分析", "sourceUrl": "https://clawhub.ai/vedantsingh60/token-watch", "canonicalUrl": "https://clawhub.ai/vedantsingh60/token-watch", "targetPlatform": "OpenClaw" }, "install": { "downloadMode": "redirect", "downloadUrl": "/downloads/token-watch", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=token-watch", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "installMethod": "Manual import", "extraction": "Extract archive", "prerequisites": [ "OpenClaw" ], "packageFormat": "ZIP package", "includedAssets": [ "LICENSE.md", "README.md", "manifest.yaml", "SKILL.md", "tokenwatch.py" ], "primaryDoc": "SKILL.md", "quickSetup": [ "Download the package from Yavira.", "Extract the archive and review SKILL.md first.", "Import or place the package into your OpenClaw setup." ], "agentAssist": { "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.", "steps": [ "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." ], "prompts": [ { "label": "New install", "body": "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." }, { "label": "Upgrade existing", "body": "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." } ] }, "sourceHealth": { "source": "tencent", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-04-23T16:43:11.935Z", "expiresAt": "2026-04-30T16:43:11.935Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=4claw-imageboard", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=4claw-imageboard", "contentDisposition": "attachment; filename=\"4claw-imageboard-1.0.1.zip\"", "redirectLocation": null, "bodySnippet": null }, "scope": "source", "summary": "Source download looks usable.", "detail": "Yavira can redirect you to the upstream package for this source.", "primaryActionLabel": "Download for OpenClaw", "primaryActionHref": "/downloads/token-watch" }, "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." ] }, "downloadPageUrl": "https://openagent3.xyz/downloads/token-watch", "agentPageUrl": "https://openagent3.xyz/skills/token-watch/agent", "manifestUrl": "https://openagent3.xyz/skills/token-watch/agent.json", "briefUrl": "https://openagent3.xyz/skills/token-watch/agent.md" }, "agentAssist": { "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.", "steps": [ "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." ], "prompts": [ { "label": "New install", "body": "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." }, { "label": "Upgrade existing", "body": "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." } ] }, "documentation": { "source": "clawhub", "primaryDoc": "SKILL.md", "sections": [ { "title": "TokenWatch", "body": "Track, analyze, and optimize token usage and costs across AI providers. Set budgets, get alerts, compare models, and reduce your spend.\n\nFree and open-source (MIT License) • Zero dependencies • Works locally • No API keys required" }, { "title": "Why This Skill?", "body": "After OpenAI's acquisition of OpenClaw, token costs are the #1 concern for power users. This skill gives you full visibility into what you're spending, where it's going, and exactly how to reduce it." }, { "title": "Problems it solves:", "body": "You don't know how much you're spending until the bill arrives\nNo way to compare costs across providers before choosing a model\nNo alerts when you're approaching your budget\nNo actionable suggestions for reducing spend" }, { "title": "1. Record Usage & Auto-Calculate Costs", "body": "from tokenwatch import TokenWatch\n\nmonitor = TokenWatch()\n\nmonitor.record_usage(\n model=\"claude-haiku-4-5-20251001\",\n input_tokens=1200,\n output_tokens=400,\n task_label=\"summarize article\"\n)\n# ✅ Recorded: $0.00192" }, { "title": "2. Auto-Record from API Responses", "body": "from tokenwatch import record_from_anthropic_response, record_from_openai_response\n\n# Anthropic\nresponse = client.messages.create(model=\"claude-haiku-4-5-20251001\", ...)\nrecord_from_anthropic_response(monitor, response, task_label=\"my task\")\n\n# OpenAI\nresponse = client.chat.completions.create(model=\"gpt-4o-mini\", ...)\nrecord_from_openai_response(monitor, response, task_label=\"my task\")" }, { "title": "3. Set Budgets with Alerts", "body": "monitor.set_budget(\n daily_usd=1.00,\n weekly_usd=5.00,\n monthly_usd=15.00,\n per_call_usd=0.10,\n alert_at_percent=80.0 # Alert at 80% of budget\n)\n# ✅ Budget set: daily=$1.0, weekly=$5.0, monthly=$15.0\n# 🚨 BUDGET ALERT fires automatically when threshold is crossed" }, { "title": "4. Dashboard", "body": "print(monitor.format_dashboard())\n\n💰 SPENDING SUMMARY\n Today: $0.0042 (4 calls, 13,600 tokens)\n Week: $0.0231 (18 calls, 67,200 tokens)\n Month: $0.1847 (92 calls, 438,000 tokens)\n\n📋 BUDGET STATUS\n Daily: [████░░░░░░░░░░░░░░░░] 42% $0.0042 / $1.00 ✅\n Monthly: [███████░░░░░░░░░░░░░] 37% $0.1847 / $0.50 ⚠️\n\n💡 OPTIMIZATION TIPS\n 🔴 Swap Opus → Sonnet for non-reasoning tasks (save ~$8.20/mo)\n 🟡 High avg cost/call on gpt-4o — reduce prompt length" }, { "title": "5. Compare Models Before Calling", "body": "# For 2000 input + 500 output tokens:\nfor m in monitor.compare_models(2000, 500)[:6]:\n print(f\"{m['model']:<42} ${m['cost_usd']:.6f}\")\n\ngemini-2.5-flash $0.000300\ngpt-4o-mini $0.000600\nmistral-small-2501 $0.000350\nclaude-haiku-4-5-20251001 $0.003600\nmistral-large-2501 $0.007000\ngemini-2.5-pro $0.007500" }, { "title": "6. Estimate Before You Call", "body": "estimate = monitor.estimate_cost(\"claude-sonnet-4-5-20250929\", input_tokens=5000, output_tokens=1000)\nprint(f\"Estimated cost: ${estimate['estimated_cost_usd']:.6f}\")" }, { "title": "7. Optimization Suggestions", "body": "suggestions = monitor.get_optimization_suggestions()\nfor s in suggestions:\n savings = s.get(\"estimated_monthly_savings_usd\", 0)\n print(f\"[{s['priority'].upper()}] {s['message']}\")\n if savings:\n print(f\" → Save ~${savings:.2f}/month\")" }, { "title": "8. Export Reports", "body": "monitor.export_report(\"monthly_report.json\", period=\"month\")" }, { "title": "Supported Models (Feb 2026)", "body": "41 models across 10 providers — updated Feb 16, 2026.\n\nProviderModelInput/1MOutput/1MAnthropicclaude-opus-4-6$5.00$25.00Anthropicclaude-opus-4-5$5.00$25.00Anthropicclaude-sonnet-4-5-20250929$3.00$15.00Anthropicclaude-haiku-4-5-20251001$1.00$5.00OpenAIgpt-5.2-pro$21.00$168.00OpenAIgpt-5.2$1.75$14.00OpenAIgpt-5$1.25$10.00OpenAIgpt-4.1$2.00$8.00OpenAIgpt-4.1-mini$0.40$1.60OpenAIgpt-4.1-nano$0.10$0.40OpenAIo3$10.00$40.00OpenAIo4-mini$1.10$4.40Googlegemini-3-pro$2.00$12.00Googlegemini-3-flash$0.50$3.00Googlegemini-2.5-pro$1.25$10.00Googlegemini-2.5-flash$0.30$2.50Googlegemini-2.5-flash-lite$0.10$0.40Googlegemini-2.0-flash$0.10$0.40Mistralmistral-large-2411$2.00$6.00Mistralmistral-medium-3$0.40$2.00Mistralmistral-small$0.10$0.30Mistralmistral-nemo$0.02$0.10Mistraldevstral-2$0.40$2.00xAIgrok-4$3.00$15.00xAIgrok-3$3.00$15.00xAIgrok-4.1-fast$0.20$0.50Kimikimi-k2.5$0.60$3.00Kimikimi-k2$0.60$2.50Kimikimi-k2-turbo$1.15$8.00Qwenqwen3.5-plus$0.11$0.44Qwenqwen3-max$0.40$1.60Qwenqwen3-vl-32b$0.91$3.64DeepSeekdeepseek-v3.2$0.14$0.28DeepSeekdeepseek-r1$0.55$2.19DeepSeekdeepseek-v3$0.27$1.10Metallama-4-maverick$0.27$0.85Metallama-4-scout$0.18$0.59Metallama-3.3-70b$0.23$0.40MiniMaxminimax-m2.5$0.30$1.20MiniMaxminimax-m1$0.43$1.93MiniMaxminimax-text-01$0.20$1.10\n\nTo add a custom model: add it to PROVIDER_PRICING dict at the top of tokenwatch.py." }, { "title": "TokenWatch(storage_path)", "body": "Initialize monitor. Data stored in .tokenwatch/ by default." }, { "title": "record_usage(model, input_tokens, output_tokens, task_label, session_id)", "body": "Record a single API call. Returns TokenUsageRecord with calculated cost." }, { "title": "set_budget(daily_usd, weekly_usd, monthly_usd, per_call_usd, alert_at_percent)", "body": "Configure spending limits. Alerts fire automatically when thresholds are crossed." }, { "title": "get_spend(period)", "body": "Get aggregated spend. Period: \"today\", \"week\", \"month\", \"all\", or \"YYYY-MM-DD\"." }, { "title": "get_spend_by_model(period)", "body": "Spending breakdown by model, sorted by cost descending." }, { "title": "get_spend_by_provider(period)", "body": "Spending breakdown by provider." }, { "title": "compare_models(input_tokens, output_tokens)", "body": "Compare costs across all known models. Returns list sorted cheapest first." }, { "title": "estimate_cost(model, input_tokens, output_tokens)", "body": "Estimate cost before making a call." }, { "title": "get_optimization_suggestions()", "body": "Analyze usage and return ranked suggestions with estimated monthly savings." }, { "title": "format_dashboard()", "body": "Human-readable spending dashboard with budget bars and tips." }, { "title": "export_report(output_file, period)", "body": "Export full report to JSON." }, { "title": "record_from_anthropic_response(monitor, response, task_label)", "body": "Helper to auto-record from Anthropic SDK response object." }, { "title": "record_from_openai_response(monitor, response, task_label)", "body": "Helper to auto-record from OpenAI SDK response object." }, { "title": "Privacy & Security", "body": "✅ Zero telemetry — No data sent anywhere\n✅ Local-only storage — Everything in .tokenwatch/ on your machine\n✅ No API keys required — The monitor itself needs no credentials\n✅ No authentication — No accounts or logins needed\n✅ Full transparency — MIT licensed, source code included" }, { "title": "[1.2.3] - 2026-02-16", "body": "📋 Updated SKILL.md model table to match code: 41 models across 10 providers" }, { "title": "[1.2.0] - 2026-02-16", "body": "✨ Added DeepSeek, Meta Llama, MiniMax providers\n✨ Expanded to 41 models across 10 providers\n✨ Updated all Anthropic/OpenAI/Google/Mistral pricing to Feb 2026 rates" }, { "title": "[1.1.0] - 2026-02-16", "body": "✨ Added xAI Grok, Kimi (Moonshot), Qwen (Alibaba)\n✨ Expanded to 32 models across 7 providers" }, { "title": "[1.0.0] - 2026-02-16", "body": "✨ Initial release — TokenWatch\n✨ Pricing table for 11 models across 5 providers\n✨ Budget alerts: daily, weekly, monthly, per-call thresholds\n✨ Model cost comparison, cost estimation, optimization suggestions\n✨ Auto-hooks for Anthropic and OpenAI response objects\n✨ Dashboard, JSON export, local-only storage, MIT licensed\n\nLast Updated: February 16, 2026\nCurrent Version: 1.2.3\nStatus: Active & Community-Maintained\n\n© 2026 UnisAI Community" } ], "body": "TokenWatch\n\nTrack, analyze, and optimize token usage and costs across AI providers. Set budgets, get alerts, compare models, and reduce your spend.\n\nFree and open-source (MIT License) • Zero dependencies • Works locally • No API keys required\n\nWhy This Skill?\n\nAfter OpenAI's acquisition of OpenClaw, token costs are the #1 concern for power users. This skill gives you full visibility into what you're spending, where it's going, and exactly how to reduce it.\n\nProblems it solves:\nYou don't know how much you're spending until the bill arrives\nNo way to compare costs across providers before choosing a model\nNo alerts when you're approaching your budget\nNo actionable suggestions for reducing spend\nFeatures\n1. Record Usage & Auto-Calculate Costs\nfrom tokenwatch import TokenWatch\n\nmonitor = TokenWatch()\n\nmonitor.record_usage(\n model=\"claude-haiku-4-5-20251001\",\n input_tokens=1200,\n output_tokens=400,\n task_label=\"summarize article\"\n)\n# ✅ Recorded: $0.00192\n\n2. Auto-Record from API Responses\nfrom tokenwatch import record_from_anthropic_response, record_from_openai_response\n\n# Anthropic\nresponse = client.messages.create(model=\"claude-haiku-4-5-20251001\", ...)\nrecord_from_anthropic_response(monitor, response, task_label=\"my task\")\n\n# OpenAI\nresponse = client.chat.completions.create(model=\"gpt-4o-mini\", ...)\nrecord_from_openai_response(monitor, response, task_label=\"my task\")\n\n3. Set Budgets with Alerts\nmonitor.set_budget(\n daily_usd=1.00,\n weekly_usd=5.00,\n monthly_usd=15.00,\n per_call_usd=0.10,\n alert_at_percent=80.0 # Alert at 80% of budget\n)\n# ✅ Budget set: daily=$1.0, weekly=$5.0, monthly=$15.0\n# 🚨 BUDGET ALERT fires automatically when threshold is crossed\n\n4. Dashboard\nprint(monitor.format_dashboard())\n\n💰 SPENDING SUMMARY\n Today: $0.0042 (4 calls, 13,600 tokens)\n Week: $0.0231 (18 calls, 67,200 tokens)\n Month: $0.1847 (92 calls, 438,000 tokens)\n\n📋 BUDGET STATUS\n Daily: [████░░░░░░░░░░░░░░░░] 42% $0.0042 / $1.00 ✅\n Monthly: [███████░░░░░░░░░░░░░] 37% $0.1847 / $0.50 ⚠️\n\n💡 OPTIMIZATION TIPS\n 🔴 Swap Opus → Sonnet for non-reasoning tasks (save ~$8.20/mo)\n 🟡 High avg cost/call on gpt-4o — reduce prompt length\n\n5. Compare Models Before Calling\n# For 2000 input + 500 output tokens:\nfor m in monitor.compare_models(2000, 500)[:6]:\n print(f\"{m['model']:<42} ${m['cost_usd']:.6f}\")\n\ngemini-2.5-flash $0.000300\ngpt-4o-mini $0.000600\nmistral-small-2501 $0.000350\nclaude-haiku-4-5-20251001 $0.003600\nmistral-large-2501 $0.007000\ngemini-2.5-pro $0.007500\n\n6. Estimate Before You Call\nestimate = monitor.estimate_cost(\"claude-sonnet-4-5-20250929\", input_tokens=5000, output_tokens=1000)\nprint(f\"Estimated cost: ${estimate['estimated_cost_usd']:.6f}\")\n\n7. Optimization Suggestions\nsuggestions = monitor.get_optimization_suggestions()\nfor s in suggestions:\n savings = s.get(\"estimated_monthly_savings_usd\", 0)\n print(f\"[{s['priority'].upper()}] {s['message']}\")\n if savings:\n print(f\" → Save ~${savings:.2f}/month\")\n\n8. Export Reports\nmonitor.export_report(\"monthly_report.json\", period=\"month\")\n\nSupported Models (Feb 2026)\n\n41 models across 10 providers — updated Feb 16, 2026.\n\nProvider\tModel\tInput/1M\tOutput/1M\nAnthropic\tclaude-opus-4-6\t$5.00\t$25.00\nAnthropic\tclaude-opus-4-5\t$5.00\t$25.00\nAnthropic\tclaude-sonnet-4-5-20250929\t$3.00\t$15.00\nAnthropic\tclaude-haiku-4-5-20251001\t$1.00\t$5.00\nOpenAI\tgpt-5.2-pro\t$21.00\t$168.00\nOpenAI\tgpt-5.2\t$1.75\t$14.00\nOpenAI\tgpt-5\t$1.25\t$10.00\nOpenAI\tgpt-4.1\t$2.00\t$8.00\nOpenAI\tgpt-4.1-mini\t$0.40\t$1.60\nOpenAI\tgpt-4.1-nano\t$0.10\t$0.40\nOpenAI\to3\t$10.00\t$40.00\nOpenAI\to4-mini\t$1.10\t$4.40\nGoogle\tgemini-3-pro\t$2.00\t$12.00\nGoogle\tgemini-3-flash\t$0.50\t$3.00\nGoogle\tgemini-2.5-pro\t$1.25\t$10.00\nGoogle\tgemini-2.5-flash\t$0.30\t$2.50\nGoogle\tgemini-2.5-flash-lite\t$0.10\t$0.40\nGoogle\tgemini-2.0-flash\t$0.10\t$0.40\nMistral\tmistral-large-2411\t$2.00\t$6.00\nMistral\tmistral-medium-3\t$0.40\t$2.00\nMistral\tmistral-small\t$0.10\t$0.30\nMistral\tmistral-nemo\t$0.02\t$0.10\nMistral\tdevstral-2\t$0.40\t$2.00\nxAI\tgrok-4\t$3.00\t$15.00\nxAI\tgrok-3\t$3.00\t$15.00\nxAI\tgrok-4.1-fast\t$0.20\t$0.50\nKimi\tkimi-k2.5\t$0.60\t$3.00\nKimi\tkimi-k2\t$0.60\t$2.50\nKimi\tkimi-k2-turbo\t$1.15\t$8.00\nQwen\tqwen3.5-plus\t$0.11\t$0.44\nQwen\tqwen3-max\t$0.40\t$1.60\nQwen\tqwen3-vl-32b\t$0.91\t$3.64\nDeepSeek\tdeepseek-v3.2\t$0.14\t$0.28\nDeepSeek\tdeepseek-r1\t$0.55\t$2.19\nDeepSeek\tdeepseek-v3\t$0.27\t$1.10\nMeta\tllama-4-maverick\t$0.27\t$0.85\nMeta\tllama-4-scout\t$0.18\t$0.59\nMeta\tllama-3.3-70b\t$0.23\t$0.40\nMiniMax\tminimax-m2.5\t$0.30\t$1.20\nMiniMax\tminimax-m1\t$0.43\t$1.93\nMiniMax\tminimax-text-01\t$0.20\t$1.10\n\nTo add a custom model: add it to PROVIDER_PRICING dict at the top of tokenwatch.py.\n\nAPI Reference\nTokenWatch(storage_path)\n\nInitialize monitor. Data stored in .tokenwatch/ by default.\n\nrecord_usage(model, input_tokens, output_tokens, task_label, session_id)\n\nRecord a single API call. Returns TokenUsageRecord with calculated cost.\n\nset_budget(daily_usd, weekly_usd, monthly_usd, per_call_usd, alert_at_percent)\n\nConfigure spending limits. Alerts fire automatically when thresholds are crossed.\n\nget_spend(period)\n\nGet aggregated spend. Period: \"today\", \"week\", \"month\", \"all\", or \"YYYY-MM-DD\".\n\nget_spend_by_model(period)\n\nSpending breakdown by model, sorted by cost descending.\n\nget_spend_by_provider(period)\n\nSpending breakdown by provider.\n\ncompare_models(input_tokens, output_tokens)\n\nCompare costs across all known models. Returns list sorted cheapest first.\n\nestimate_cost(model, input_tokens, output_tokens)\n\nEstimate cost before making a call.\n\nget_optimization_suggestions()\n\nAnalyze usage and return ranked suggestions with estimated monthly savings.\n\nformat_dashboard()\n\nHuman-readable spending dashboard with budget bars and tips.\n\nexport_report(output_file, period)\n\nExport full report to JSON.\n\nrecord_from_anthropic_response(monitor, response, task_label)\n\nHelper to auto-record from Anthropic SDK response object.\n\nrecord_from_openai_response(monitor, response, task_label)\n\nHelper to auto-record from OpenAI SDK response object.\n\nPrivacy & Security\n✅ Zero telemetry — No data sent anywhere\n✅ Local-only storage — Everything in .tokenwatch/ on your machine\n✅ No API keys required — The monitor itself needs no credentials\n✅ No authentication — No accounts or logins needed\n✅ Full transparency — MIT licensed, source code included\nChangelog\n[1.2.3] - 2026-02-16\n📋 Updated SKILL.md model table to match code: 41 models across 10 providers\n[1.2.0] - 2026-02-16\n✨ Added DeepSeek, Meta Llama, MiniMax providers\n✨ Expanded to 41 models across 10 providers\n✨ Updated all Anthropic/OpenAI/Google/Mistral pricing to Feb 2026 rates\n[1.1.0] - 2026-02-16\n✨ Added xAI Grok, Kimi (Moonshot), Qwen (Alibaba)\n✨ Expanded to 32 models across 7 providers\n[1.0.0] - 2026-02-16\n✨ Initial release — TokenWatch\n✨ Pricing table for 11 models across 5 providers\n✨ Budget alerts: daily, weekly, monthly, per-call thresholds\n✨ Model cost comparison, cost estimation, optimization suggestions\n✨ Auto-hooks for Anthropic and OpenAI response objects\n✨ Dashboard, JSON export, local-only storage, MIT licensed\n\nLast Updated: February 16, 2026 Current Version: 1.2.3 Status: Active & Community-Maintained\n\n© 2026 UnisAI Community" }, "trust": { "sourceLabel": "tencent", "provenanceUrl": "https://clawhub.ai/vedantsingh60/token-watch", "publisherUrl": "https://clawhub.ai/vedantsingh60/token-watch", "owner": "vedantsingh60", "version": "1.2.3", "license": null, "verificationStatus": "Indexed source record" }, "links": { "detailUrl": "https://openagent3.xyz/skills/token-watch", "downloadUrl": "https://openagent3.xyz/downloads/token-watch", "agentUrl": "https://openagent3.xyz/skills/token-watch/agent", "manifestUrl": "https://openagent3.xyz/skills/token-watch/agent.json", "briefUrl": "https://openagent3.xyz/skills/token-watch/agent.md" } }