Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    llm cost optimization audit

    1

    Audit LLM spend to find recoverable waste in prompt caching, model routing, and token usage.

    $11

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    llm cost optimization audit

    llm cost optimization audit

    Example session with this skill installed

    Our Claude bill is spiking. We added prompt caching but it hasn't moved the needle. Here is our prompt template and a sample of our usage logs. Can you find the leak?

    • Read your context and instructions
    • Compiled the llm cost optimization

    The audit shows cache_read_input_tokens is zero despite cache_control being set. Your prompt template includes a 'current_time' variable at the start of the system prompt. This invalidates the cache for every request. Moving this after the stable prefix will save an estimated 70% on input tokens.

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

    About this skill

    The problem

    LLM bills often grow into unmanaged line items that eat product margins. Developers frequently mistake setting a configuration flag for actually saving money, leading to "optimized" prompts that are silently more expensive than the baseline.

    What it does

    • Profiles token usage across input, output, and cache creation to find the real cost drivers.
    • Identifies silent cache invalidators like per-request timestamps or unsorted JSON that break prompt caching.
    • Calculates the break-even point for caching based on reuse frequency and TTL windows.
    • Audits model routing to find high-volume, low-complexity tasks running on over-powered tiers.
    • Sizes estimated savings for request batching and context compaction before you write any code.

    Frameworks & tools

    Works with any LLM provider usage logs. Specific logic for Anthropic Claude API prompt caching and Batch API. Uses current pricing snapshots for ROI calculations.

    Why this beats prompting it yourself

    Generic prompts often guess at savings. This skill uses a "profile first" discipline to ensure recommendations are based on your actual token distribution. It accounts for hidden costs like the prompt-cache write premium, which can actually increase your bill if reuse is infrequent.

    Use cases

    • Diagnosing why prompt caching is resulting in zero read tokens despite being enabled.
    • Evaluating the margin impact of moving high-latency background jobs to a Batches API.
    • Reducing context waste in long-running agent loops or chat histories.
    • Right-sizing models for specific tasks like classification or extraction without losing quality.

    Known limitations

    Does not perform quality benchmarks or "capability bake-offs" between models. Focuses on inference costs only, excluding database, hosting, or egress fees.

    Does not perform capability bake-offs or quality evals. Limited to inference cost auditing. Requires user to provide usage data or architecture descriptions for accuracy.

    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 3 months ago

    Permissions

    No special permissions declared or detected

    Listed3 months ago

    What's inside

    Frequently Asked Questions

    Expect a detailed breakdown of your token distribution, specific identification of 'cache-breaking' patterns in your prompts, and a tiered list of savings opportunities with projected ROI for each.

    While it offers specialized logic for Anthropic's prompt caching and Batch API, the skill is provider-agnostic and works with any LLM usage logs that include input, output, and cache metadata.

    No, this skill focuses strictly on inference costs; it identifies areas to reduce spend without running benchmark tests or 'evals' to measure model performance impacts.

    Simply provide your LLM provider's usage logs or export files; the skill is designed to ingest standard JSON or CSV usage data to perform its analysis.

    This audit identifies 'silent invalidators'—like dynamic timestamps or non-deterministic JSON—that may prevent your prompts from being cached despite having the feature enabled.

    Pricing is updated via regular snapshots to reflect the latest model costs, ensuring your savings projections are accurate to current market rates.

    New in Agents & Orchestration