cost-report

Generate weekly API cost reports from token usage logs with per-skill and per-model breakdowns.

6|2|Updated May 21, 2026

Search Tags:#token usage #anomaly detection #budget monitoring #cost forecasting #model pricing #api costs #optimization recommendations

Ready to run

npx skills add https://github.com/anajuliabit/aeon --skill cost-report-anajuliabit

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: cost-report
Source: https://github.com/anajuliabit/aeon/tree/main/skills/cost-report
Command: npx skills add https://github.com/anajuliabit/aeon --skill cost-report-anajuliabit

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It turns raw token-usage logs into a weekly, decision-ready cost report so you can spot anomalies, predict burn, and immediately act on concrete optimization opportunities.

Core Features & Use Cases

Weekly cost breakdown: Calculates total spend across input/output/cache read/write and ranks top skills by cost.
Anomaly + week-over-week detection: Flags unusually expensive runs and identifies cost shifts versus the prior window.
Actionable optimization plan: Recommends up to three specific changes (e.g., model downgrade, cache usage improvements, or config drift) with estimated dollar savings.
Burn forecast: Projects 30-day spend from the observed daily average and calls out high burn-rate risk.

Quick Start

Use the cost-report skill to generate a cost report covering the last 7 days by running it with N="7" days.

Frequently Asked Questions about cost-report

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I generate a weekly API cost report from token usage logs?▼

You can generate a weekly API cost report by reading token usage logs from a CSV file and applying model pricing tables to convert usage into dollar spend. This provides per-skill, per-model, and week-over-week cost breakdowns for monitoring.

What is anomaly detection in API costs and how does it work?▼

Anomaly detection in API costs identifies unusually expensive runs and flags cost shifts compared to the prior time window. It works by analyzing token usage logs to detect sudden spend increases, helping you catch unexpected model behavior or configuration drift.

Can I forecast budget burn for scheduled autonomous agent runs?▼

Yes, you can forecast budget burn for scheduled autonomous agent runs by projecting 30-day spend from the observed daily average token usage. This calls out high burn-rate risk to help you adjust workloads before exceeding your budget.

How do I calculate model pricing and API costs from token usage?▼

You calculate model pricing and API costs from token usage by applying pricing tables to input, output, cache read, and cache write token counts. This computes total dollar spend across different models and skills for accurate budget monitoring.

What happens if my token usage CSV data is missing when monitoring costs?▼

When token usage CSV data is missing during cost monitoring, the report generation safely skips the process. Defensive CSV parsing ensures the tool handles missing data gracefully without crashing or producing inaccurate cost forecasts.

Why does my API cost optimization plan recommend model downgrades and cache usage improvements?▼

An API cost optimization plan recommends model downgrades and cache usage improvements because they offer concrete dollar savings. Analyzing token usage logs identifies where cheaper models or better caching can reduce spend without sacrificing functionality.