Analyze specialist KPI data from observability SQLite databases.
npx skills add https://github.com/xtrm-dev/specialists --skill using-kpi
Please help me install this Agent Skill. Skill: using-kpi Source: https://github.com/xtrm-dev/specialists/tree/main/config/skills/using-kpi Command: npx skills add https://github.com/xtrm-dev/specialists --skill using-kpi
This Skill helps you analyze specialist KPI data from observability SQLite so you can understand runtime cost, payload growth, waiting behavior, and outliers without manually stitching together queries.
Ask the skill to analyze specialist KPI data from sp db stats or sp db extract and return the most important performance, waiting, payload, and outlier findings.
High-intent search queries and answers about installing and using this skill.
To analyze specialist KPI data from SQLite observability databases, you can query sp db stats or sp db extract workflows to rank active runtime, detect outliers, and inspect tool-call distributions for performance tuning.
Detecting runtime outliers and payload bloat involves identifying jobs above the p95 threshold for runtime, tool calls, turns, and payload size using SQLite query patterns applied to your observability database.
Yes, you can compare specialist and model performance by generating a leaderboard that ranks combinations by active runtime, total runtime, turns, tools, and payload size extracted from SQLite.
Inspecting waiting-state issues and payload-to-runtime correlation requires applying SQLite query patterns to observability data to evaluate operational hygiene and identify areas needing prompt or workflow optimization.
Yes, specialist KPI runtime analysis requires cl100k_base-style token estimation alongside active runtime ranking and SQLite query patterns to accurately evaluate leaderboard, hygiene, and component breakdown metrics.
Use SQLite query patterns for specialist KPI performance tuning when you need to understand runtime cost, payload growth, and waiting behavior without manually stitching together complex database queries.
daudaudinangValidate Vibecode runtime health and report READY or DEGRADED status.
brandonburrusApplies statistical methods to analyze data distributions, detect anomalies, and test hypotheses.
payloadcmsAutomate Payload CMS UI reskinning from Figma designs with SCSS to CSS migration.
jmolzAnalyze data distributions, trends, outliers, and hypothesis tests using Python.
