Detect anomalous per-unit rates in ReturnPro financial records using z-scores.
npx skills add https://github.com/clenisa/optimal-cli --skill rate-anomalies
Please help me install this Agent Skill. Skill: rate-anomalies Source: https://github.com/clenisa/optimal-cli/tree/main/skills/rate-anomalies Command: npx skills add https://github.com/clenisa/optimal-cli --skill rate-anomalies
Detects anomalous per-unit rates across program-month entries in ReturnPro staged financial data to surface data entry errors, misclassified programs, and upstream reporting issues that distort financial KPIs and automated analytics.
Run the rate-anomalies skill against recent ReturnPro staged financials to flag and export program-month $/unit outliers.
High-intent search queries and answers about installing and using this skill.
Detect anomalous per-unit rates by computing rate z-scores for each program-month and flagging statistical outliers that exceed a configurable threshold. This process calculates rate as revenue divided by units and compares it to the historical mean and standard deviation per program.
The best way to find $/unit rate outliers is to run anomaly detection against recent staged financials, filtering by selectable months and program names. It flags rate spikes indicating data entry mistakes or misattributed revenue streams for triage and correction.
Z-score anomaly detection works by computing the historical mean and standard deviation of per-unit rates per program. It then calculates a z-score for each program-month entry, flagging records that exceed a statistical threshold as outliers requiring data quality review.
Yes, you need Supabase credentials to analyze ReturnPro staged financials. The process requires RETURNPRO_SUPABASE_URL and RETURNPRO_SUPABASE_SERVICE_KEY environment variables to access and query the stg_financials_raw and dim_master_program tables.
Yes, you can filter rate anomaly detection by selectable months and program name substrings. This allows you to target specific time ranges and programs, sorting flagged anomalies by severity to support reporting corrections and downstream analytics.
Financial KPIs are distorted when misclassified program revenue creates anomalous per-unit rate outliers. Detecting these statistical outliers through z-score analysis surfaces upstream reporting issues and data entry errors, allowing corrections before they impact automated analytics.
clenisaUpsert aggregated R1 XLSX data into ReturnPro's stg_financials_raw.
UnicornAnalyze tabular CSV, JSON, and spreadsheet datasets to produce statistical summaries, anomalies, and visualizations.
clenisaCompare staged NetSuite financials to confirmed income statements and report per-month accuracy.
anton-abyzovDetect anomalies and outliers using statistical methods, Isolation Forest, and autoencoders.
