Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    AI Agency Client ROI & Unit-Cost Reporter

    1

    Convert AI agent usage and outcome CSV/JSON exports into transparent API unit-cost metrics and a printable HTML report. Identify unknown costs, duplicate calls and incomplete data without inventing financial ROI.

    $9.99

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    AI Agency Client ROI & Unit-Cost Reporter

    AI Agency Client ROI & Unit-Cost Reporter

    Example session with this skill installed

    Analyze the synthetic AI agent usage records in examples/synthetic-usage.csv and the business outcomes in examples/synthetic-outcomes.csv using AI Agency Client ROI & Unit-Cost Reporter v1.0.4.

    Calculate total AI API costs, token usage, cost per conversation and cost per confirmed business outcome.

    Check for duplicate records, missing costs and incomplete data. Clearly distinguish buyer-supplied costs from estimates.

    Generate an evidence-based financial report in text format. Do not invent revenue, profit or business ROI.

    • Read your context and instructions
    • Compiled the ai agency client

    AI Agency Client ROI & Unit-Cost Reporter v1.0.4

    Evidence coverage: COMPLETE_FOR_SUPPLIED_DATA

    Conversations: 6; unique API calls: 8

    Input tokens: 177000; cached input: 9000; output tokens: 15400

    AI API cost (USD): 0.46 (complete for supplied calls, not provider-verified)

    Cost provenance: Buyer-supplied per-call API costs; not independently reconciled against a provider invoice.

    Cost source counts: buyer-supplied=8; rate-derived estimates=0; unpriced=0

    Average AI cost per conversation (USD): 0.08

    Reported business outcomes (buyer-supplied):

    • booked: 2
    • handoff: 1
    • qualified: 1
    • resolved: 1
    • unresolved: 1

    All-workload AI spend per confirmed outcome:

    • cost per booked: USD 0.23
    • cost per qualified: USD 0.46
    • cost per resolved: USD 0.46

    Per-conversation breakdown (hashed IDs; incomplete if noted):

    Rounding note: Total cost is calculated from unrounded call amounts. Per-conversation values are independently rounded to cents and may not add to the rounded total.

    • e8de016fbd70 | calls=2 | input=65000 | cached_input=5000 | output=7000 | USD 0.18 | unresolved | basis=buyer-supplied amount
    • 856b6b7faff0 | calls=1 | input=60000 | cached_input=0 | output=4500 | USD 0.16 | handoff | basis=buyer-supplied amount
    • 18824f024f41 | calls=1 | input=22000 | cached_input=0 | output=2000 | USD 0.06 | booked | basis=buyer-supplied amount
    • fa57a52dbf08 | calls=2 | input=15000 | cached_input=1000 | output=1400 | USD 0.04 | booked | basis=buyer-supplied amount
    • 9e518076ab40 | calls=1 | input=12000 | cached_input=3000 | output=300 | USD 0.02 | qualified | basis=buyer-supplied amount
    • 219885bcf6b7 | calls=1 | input=3000 | cached_input=0 | output=200 | USD 0.01 | resolved | basis=buyer-supplied amount

    Limitations

    • No connection to provider APIs, WhatsApp, CRM or billing systems.
    • Prices and usage come only from buyer supplied records; no built-in current model tariff claims.
    • Business unit cost = all observed workload AI spend / count of confirmed outcomes, not just successful-session spend.
    • Does not measure WhatsApp fees, automation software, hosting, taxes or staff time.
    • An outcome label is supplied by the buyer and is not authenticated by this script.
    • Use anonymized conversation IDs; never upload customer PII or credentials.

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

    About this skill

    Understand what your AI agents actually cost, and explain it clearly to your clients.

    AI Agency Client ROI & Unit-Cost Reporter helps automation agencies, AI developers and consultants transform technical usage records into clear financial reports for business owners.

    Instead of presenting confusing token counts or unsupported return-on-investment claims, this skill provides evidence-based AI operating cost metrics, with transparent calculations and explicit data limitations.

    What it calculates:

    • Total AI API calls and conversations using stable identifiers.
    • Input, cached-input and output token consumption.
    • AI API costs from buyer-provided records.
    • Estimated costs using explicitly supplied token pricing.
    • Average AI API cost per conversation when sufficient evidence is available.
    • AI API cost per buyer-labeled booking, qualified lead or resolved conversation.
    • Per-conversation usage and cost breakdowns.
    • Cost coverage and data completeness indicators.

    Built-in data quality protections:

    • Detects and deduplicates identical API call records.
    • Rejects conflicting duplicate records, including conflicting cost sources.
    • Identifies missing conversation identifiers.
    • Distinguishes unknown costs from verified zero values.
    • Prevents incomplete data from producing misleading averages.
    • Separates buyer-supplied costs from rate-derived estimates.
    • Handles monetary rounding transparently.
    • Rejects malformed token counts and invalid monetary values.
    • Uses pseudonymized identifiers in reports.

    What you receive:

    A structured financial analysis available in text, JSON and printable HTML formats.

    Reports include cost summaries, token usage, business outcome counts, unit-cost indicators, evidence coverage, calculation notes and limitations.

    The package includes a deterministic Python analysis script, data contracts, reporting templates, synthetic examples and automated tests.

    How it works:

    1. Export your AI agent usage records in CSV or JSON format.
    2. Provide matching conversation outcomes when available.
    3. Optionally supply your model pricing information.
    4. Run the included offline Python checker.
    5. Review the generated financial report and its evidence limitations.

    No external API connections, paid services or proprietary infrastructure are required.

    Why it matters:

    Agencies often need to explain AI operating costs to clients who do not understand tokens, model pricing or API usage.

    This skill helps translate those records into practical unit-cost information while avoiding unsupported financial claims.

    Important limitations:

    Despite its name, this tool does not calculate actual business ROI, revenue, profit or financial returns.

    It measures AI API costs based on buyer-provided records. Pricing-based calculations are labeled as estimates, and supplied costs are not independently verified against provider invoices.

    Business outcomes are supplied by the buyer and are not automatically confirmed against CRM or booking systems.

    WhatsApp charges, automation subscriptions, hosting, staff expenses, taxes and other operational costs are excluded.

    Incomplete or inconsistent data may prevent reliable unit-cost calculations.

    The skill does not connect to production systems, retrieve credentials or modify business records.

    Compatibility:

    Designed for AI coding agents supporting SKILL.md instructions.

    Includes an offline Python 3.9+ script using only the standard library. Independently tested with OpenAI Codex on Windows 11.

    Built by Tecérale AI Skills.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Research & Analysis.

    This skill analyzes buyer-supplied CSV and JSON records. It does not connect to live LLM providers, CRM systems, WhatsApp, billing platforms or external APIs.

    It calculates AI API usage costs and unit-cost indicators, not actual business ROI, revenue, profit or investment returns.

    Buyer-supplied costs and business outcomes are not independently verified. Costs calculated from supplied token rates are explicitly labeled as estimates.

    Missing prices, incomplete outcomes, unmatched records or unreliable conversation identifiers may prevent accurate unit-cost calculations. Conflicting duplicate records are rejected.

    The analysis excludes WhatsApp fees, hosting, automation subscriptions, staff expenses, taxes and other operational costs.

    Python 3.9+ is optional for running the included offline analysis script. Reports depend on the quality and completeness of the supplied records. The skill does not modify production systems or retrieve private account data.

    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 buying a skill.

    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 6 days ago

    Nothing to install. It runs in your AI agent.

    Optional
    Python 3.9 or later

    Python 3.9+ is optional for running the local financial analysis script. Provide sanitized CSV or JSON usage records, with conversation outcomes and pricing data when available. Reports depend on the completeness of supplied data and do not calculate actual business ROI.

    Permissions

    Terminal / Shell
    Read Files

    File Scopes

    **/*.csv
    **/*.json

    Terminal access runs the included offline Python reporting script. Read Files access is limited to authorized CSV/JSON usage records, outcome data, optional pricing files and skill documentation. The checker does not require network access, credentials or automatic file modifications. Saving an HTML report requires separate user approval.

    Listed6 days ago

    What's inside

    Frequently Asked Questions

    The skill converts technical AI usage records into understandable financial indicators, including total API spending, cost per conversation and cost per buyer-reported booking, qualified lead or resolved conversation. It generates structured reports in text, JSON and printable HTML formats, helping agencies explain AI operating costs to business owners. All calculations are evidence-based, with clear warnings about missing information, estimates and rounding differences. It does not calculate actual business ROI or guarantee that an AI agent generates profit.

    The skill is provider-agnostic and analyzes standardized CSV or JSON records supplied by the buyer. Usage data may originate from different LLM providers or agent frameworks, provided it follows the documented input contract and contains sufficient identifiers, token usage and cost information. It does not automatically connect to OpenAI, Anthropic, Google Gemini, n8n or other platforms. Provider-specific exports may require conversion into the supported format. The skill was independently tested with OpenAI Codex on Windows 11. Compatibility with other AI coding agents has not been independently verified.

    No. The analysis runs locally using sanitized CSV or JSON records that you provide. There is no need to share API keys, passwords, production credentials or customer personal information. The included Python checker uses only the standard library and does not require internet access, paid APIs or external infrastructure. Business outcomes such as bookings and qualified leads must be supplied separately when those indicators are required.

    The one-time purchase includes the complete AI Agency Client ROI & Unit-Cost Reporter skill package. You receive SKILL.md instructions, an offline Python analysis script, documented CSV/JSON input contracts, synthetic examples, automated tests, reporting guidelines and reusable report templates. The script supports text, JSON and printable HTML financial reports. No recurring Tecérale subscription or paid external service is required to run the included local checker. The purchase does not include hosted dashboards, live integrations, automatic data collection or financial consulting services.

    The skill uses strict data validation to avoid misleading financial conclusions. API calls without valid conversation identifiers are rejected rather than assigned to an invented conversation. Missing costs are marked as unknown, not zero. When cost information is incomplete, the report identifies the known subtotal and avoids presenting it as the full bill. Missing or unmatched business outcomes prevent the calculation of unit-cost indicators that require complete outcome coverage. Conflicting duplicate records are rejected, while genuinely identical records are counted only once. The report clearly distinguishes complete, partial and unverifiable information so buyers can identify which conclusions are supported by their data.

    New in Research & Analysis