More screenshots
Works with the AI Agents you already use
LLM Usage Cost Allocation Pack
Allocate supplied LLM usage costs to clients and cost centers, isolate unmapped rows, compare an invoice total and flag budget overruns—without account access.
$5
LLM Usage Cost Allocation Pack
Example session with this skill installed
Use the included sample usage, mapping, invoice and budget files to prepare a local LLM cost-allocation review packet. Show totals, unmapped rows and any budget overrun. Do not access a provider account or estimate model prices.
- Read your context and instructions
- Compiled the llm usage cost
- Generated the data export
LLM Usage Cost Allocation — included sample
Status
REVIEW. The local run processed 3 supplied usage rows totaling USD 5.830000. Two rows were allocated across two cost centers for USD 5.550000; one row remains unallocated or ambiguous. The supplied invoice total is USD 5.830000, so the example has no invoice-total discrepancy.
| Cost center | Client | Allocated cost | Budget status |
|---|---|---|---|
| CLIENT-A-GENERAL | Client Alpha / General | USD 1.350000 | WITHIN_OR_NOT_SUPPLIED |
| CLIENT-A-RAG | Client Alpha / RAG | USD 4.200000 | OVER by USD 0.200000 |
The packet contains result.json, allocation.csv, unallocated.csv, USAGE_ALLOCATION.md and evidence-receipt.json. The receipt binds supplied inputs and generated outputs with SHA-256. Review the unallocated row and the RAG budget overrun before using the allocation for chargeback. This is evidence from the buyer-supplied files; it does not fetch accounts, calculate provider prices or make accounting decisions.
allocation.csv
CSV · data export
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
This downloadable local package turns authorized JSON or JSONL usage evidence into a fixed, reproducible cost-allocation packet. It includes an installable Python CLI, a local stdio MCP adapter and a companion Agent Skill. One run creates result.json, allocation.csv, unallocated.csv, USAGE_ALLOCATION.md and an evidence-receipt.json with SHA-256 bindings. It allocates supplied cost_usd values using buyer-provided mappings, isolates ambiguous rows, compares an invoice total and flags supplied budget overruns. The core is deterministic, rejects unsafe paths and malformed JSON, never overwrites an existing packet and escapes formula-active CSV cells. Processing stays local: no API key, login, upload, telemetry or live provider action is included. Useful for AI consultants preparing client billing evidence, small teams doing chargeback and operators comparing a provider export to one invoice observation. Requires Windows 11, Python 3.11–3.14 and local files in the documented schemas; optional MCP adapter uses mcp==2.2.0. Limitations: it does not fetch provider data, retrieve current model prices or provide tax or accounting advice. Costs come only from supplied values.
Costs use only supplied cost_usd values. No live provider fetch, current model pricing, tax advice or accounting decision. Requires buyer-supplied local files in documented schemas.
How to use
- 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
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 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 2 months ago
Requires
Permissions
Runs a local Python CLI or optional stdio MCP adapter; reads only buyer-selected usage, mapping, budget and invoice evidence; writes a new output packet. No browser, network or environment-variable access is required.