More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Weekly FinOps Token Budget — Multi-Model Cost Tracker

    1

    Automated weekly token ledger tracking task attribution, multi-model splits (Flash vs Pro), and local vs cloud compute.

    $25

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Weekly FinOps Token Budget — Multi-Model Cost Tracker

    Weekly FinOps Token Budget — Multi-Model Cost Tracker

    Example session with this skill installed

    Generate a weekly FinOps token report for Week 42. Break down tokens by task, model used, server vs local Mac compute split, and memory pressure.

    • Read your context and instructions
    • Compiled the weekly finops token

    Weekly FinOps Token & Compute Report (Week 42)

    • Period: 2026-10-18 to 2026-10-25
    • Total Recorded Tasks: 38 tasks
    • Weekly Token Volume: 190,800 tokens (19.1% of 1,000,000 budget ceiling)
    • Estimated Spend: ~$0.28 USD
    Model ArchitectureTasks RunInput/Output TokensEstimated CostShare
    Gemini 3.8 Flash32 tasks142,000 tok.$0.0274.4%
    Claude Sonnet 3.76 tasks48,800 tok.$0.2625.6%
    TOTAL SPRINT ROLLUP38 tasks190,800 tok.$0.28100%

    Compute Infrastructure Split:

    • ☁️ Cloud API Server: 36s (42% compute latency)
    • 💻 Local Workstation (AST, I/O, Git, FFmpeg): 50s (58% compute duration)

    Next Steps:

    Sprint burn rate is optimal (19.1%). Archive ledger snapshot to cloud backup if needed.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Weekly FinOps Token Budget — Multi-Model Cost Tracker

    Automated weekly token ledger tracking task attribution, multi-model splits (Flash vs Pro), and local vs cloud compute across 7-day rolling sprints.

    ⚡ Key Features & Engineering Guardrails:

    • Deterministic Task Attribution: Logs every completed LLM task turn with exact token volume, active model tier, and compute duration in sub-millisecond local RAM.
    • Multi-Model Cost Economics: Automatically calculates blended USD API spend per model (Gemini Flash, Claude Sonnet, GPT-4o, DeepSeek) using real-time pricing formulas.
    • Cloud vs Local Compute Split: Pinpoints engineering bottlenecks by isolating cloud API inference latency from local workstation I/O, Git operations, and asset builds.
    • Sprint Budget Ceiling & Tripwires (--budget): Configurable weekly token ceilings with a 75% early warning threshold and a 100% hard ceiling to prevent runaway reasoning loops.
    • 100% Air-Gapped & Zero Telemetry: Local JSON database stored in user directory. Zero external analytics or cloud leakage.

    📦 What's Included in the ZIP:

    • scripts/token_tracker.py — Production zero-dependency CLI tracker & cost attribution engine
    • references/finops_token_management.md — Enterprise specification for model tiering and token economics
    • SKILL.md — Lean Core integration rules (< 40 lines) for coding agents
    • Automated weekly sprint rotation and Markdown report generators

    Requires manual logging via CLI scripts. Does not pull live billing data from cloud provider consoles; attribution is based on local script execution.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 11 days ago

    Requires

    python3
    macOS
    Linux
    Windows

    Permissions

    Terminal / Shell
    Read Files
    Listed29 days ago
    Updated11 days ago

    What's inside

    Frequently Asked Questions

    Unlike standard dashboards, this skill captures local compute time alongside API tokens, providing a unified ledger for hybrid workflows that include both local workstations and cloud models.

    The skill consists of the `token_tracker.py` engine, a configuration template for setting budget tripwires, and documentation for integrating the logging commands into your existing Python-based AI workflows.

    This is a zero-dependency Python tool, making it compatible with any environment that runs Python 3.x, including local terminals, CI/CD pipelines, and various AI agent frameworks.

    No, the tool uses a deterministic logging protocol that relies on the parameters you pass during execution; it does not require your private API keys to calculate costs.

    While the skill requires script-based logging, it is designed to be easily integrated into your 'main' execution loops, allowing you to automate budget alerts and report generation as part of your development pipeline.

    New in LLM Ops