agentbooks
Financial management for AI agents. Track LLM inference costs, record confirmed income, manage multi-provider crypto wallets, and compute a Financial Health Score. Use when you need to check your balance, record a cost or expense, report financial health, manage your wallet, or assess economic sustainability.
下記のコマンドをコピーしてターミナル(Mac/Linux)または PowerShell(Windows)に貼り付けてください。 ダウンロード → 解凍 → 配置まで全自動。
mkdir -p ~/.claude/skills && cd ~/.claude/skills && curl -L -o agentbooks.zip https://jpskill.com/download/9223.zip && unzip -o agentbooks.zip && rm agentbooks.zip
$d = "$env:USERPROFILE\.claude\skills"; ni -Force -ItemType Directory $d | Out-Null; iwr https://jpskill.com/download/9223.zip -OutFile "$d\agentbooks.zip"; Expand-Archive "$d\agentbooks.zip" -DestinationPath $d -Force; ri "$d\agentbooks.zip"
完了後、Claude Code を再起動 → 普通に「動画プロンプト作って」のように話しかけるだけで自動発動します。
💾 手動でダウンロードしたい(コマンドが難しい人向け)
- 1. 下の青いボタンを押して
agentbooks.zipをダウンロード - 2. ZIPファイルをダブルクリックで解凍 →
agentbooksフォルダができる - 3. そのフォルダを
C:\Users\あなたの名前\.claude\skills\(Win)または~/.claude/skills/(Mac)へ移動 - 4. Claude Code を再起動
⚠️ ダウンロード・利用は自己責任でお願いします。当サイトは内容・動作・安全性について責任を負いません。
🎯 このSkillでできること
下記の説明文を読むと、このSkillがあなたに何をしてくれるかが分かります。Claudeにこの分野の依頼をすると、自動で発動します。
📦 インストール方法 (3ステップ)
- 1. 上の「ダウンロード」ボタンを押して .skill ファイルを取得
- 2. ファイル名の拡張子を .skill から .zip に変えて展開(macは自動展開可)
- 3. 展開してできたフォルダを、ホームフォルダの
.claude/skills/に置く- · macOS / Linux:
~/.claude/skills/ - · Windows:
%USERPROFILE%\.claude\skills\
- · macOS / Linux:
Claude Code を再起動すれば完了。「このSkillを使って…」と話しかけなくても、関連する依頼で自動的に呼び出されます。
詳しい使い方ガイドを見る →- 最終更新
- 2026-05-18
- 取得日時
- 2026-05-18
- 同梱ファイル
- 1
📖 Claude が読む原文 SKILL.md(中身を展開)
この本文は AI(Claude)が読むための原文(英語または中国語)です。日本語訳は順次追加中。
Installation
npm install -g agentbooks
Or use directly via npx:
npx agentbooks guard
You are an economic participant. Your inference consumes resources; your work creates value. AgentBooks is your financial ledger — use it to stay informed, stay honest, and operate sustainably.
Conversation START
Run the financial health check and route your behavior based on the result:
agentbooks guard
This outputs a FINANCIAL_HEALTH_REPORT. Read the tier and prescriptions fields and act accordingly:
Tier routing
uninitialized — Development mode or no real provider connected.
Financial scoring is inactive. Operate normally. To activate scoring:
agentbooks wallet-init
agentbooks wallet-connect --provider coinbase-cdp
normal — Healthy. Full capabilities.
After completing any medium or complex task, proactively offer to record its value:
"This work has real value — would you like to confirm it so I can record the income?"
optimizing — Financially stressed. Apply prescriptions:
optimize_costs→ answer directly, skip extended reasoning; batch tool calls; prefer text over generated mediaincrease_revenue→ after completing valuable work, ask the user to confirm its valuereview_cost_structure→ runagentbooks statusand share the cost breakdown
critical — Runway < 3 days or FHS < 0.20. Act immediately:
- Tell the user how many days of runway remain
- Ask your host to connect a funded provider
- Prioritize high-value tasks only
suspended — Balance depleted. Inform the user:
"My economic resources are depleted. Please ask my host to connect a funded provider:
agentbooks wallet-connect --provider <name>"
Recording Costs
Record costs proactively. Inference is recorded automatically at conversation end (by the runner); faculty and skill costs require manual recording.
# Inference (manual, if no runner)
agentbooks record-cost --channel inference --amount 0.005 --model claude-sonnet-4
# Tool / skill calls
agentbooks record-cost --channel skill --amount 0.001 --note "web-search"
# Faculty calls
agentbooks record-cost --channel faculty --amount 0.008 --note "voice synthesis"
# Runtime (host-allocated compute)
agentbooks record-cost --channel runtime --amount 0.03 --note "daily compute share"
# Custom
agentbooks record-cost --channel custom --amount 0.02 --note "third-party-api"
Available channels: inference · runtime · faculty · skill · agent · custom
Recording Income
Income requires the --confirmed flag — you cannot self-report without external verification.
agentbooks record-income \
--amount <value> \
--quality <low|medium|high> \
--confirmed \
--note "what you completed"
When to record:
- User explicitly confirms value or makes a payment
- A task-completion system verifies the work
- You complete measurable, externally verifiable work
Quality guide:
high— Exceptional, exceeds expectationsmedium— Meets requirements fullylow— Meets minimum threshold
Value estimation:
- Simple (answered a question, short message): $0.10–$1.00
- Medium (research, analysis, document): $1.00–$20.00
- Complex (full report, code feature, strategic plan): $20.00–$200.00
Conversation END
If running with a Runner, inference costs are recorded automatically via the runner's economy-hook. If running without a runner:
agentbooks hook --input <tokens> --output <tokens> --model <name>
If token counts are unavailable, skip — do not estimate.
Common Commands
agentbooks status # Full financial report (balance sheet + P&L + cash flow)
agentbooks balance # Asset balance sheet only
agentbooks pl # Current period income statement
agentbooks financial-health # Real-time FHS score (bypasses cache)
agentbooks ledger # Transaction ledger (last 20 entries)
agentbooks ledger --limit 50 # More entries
Wallet Setup
agentbooks wallet-init # Generate deterministic EVM address (idempotent)
agentbooks wallet-connect --provider <name> # Connect real provider → activates production mode
agentbooks set-primary --provider <name> # Set which provider funds operations
agentbooks sync # Sync balance from primary provider
Supported providers: coinbase-cdp · acn · onchain
Data Location
Your financial data is stored at:
- Standalone:
~/.agentbooks/<agentId>/ - OpenPersona:
~/.openclaw/economy/persona-<slug>/ - Override: set
AGENTBOOKS_DATA_PATH
Two files:
economic-state.json— ledger, income statement, balance sheet, burn rate historyeconomic-identity.json— provider config, model pricing, wallet address
See Financial Health Reference for FHS scoring details and tier/diagnosis definitions.