NubaeonNubaeonCommunityยท1 Agent Skills Included

empirica

Measure, gate, and remember what coding agents know

Measures what an AI coding agent actually knows before it edits code, gating changes until understanding is demonstrated. Persists findings, dead-ends, and decisions across sessions so agents stop repeating the same questions and mistakes. Includes code audits, architecture reviews, quality sweeps, and calibration reports that turn agent reliability into trackable data.
npx skills add Nubaeon/empirica --all -g -y
Available:

Instructs the AI agent to track all work through the bd issue system and Empirica goals, with a step-by-step workflow for claiming, completing, and committing tasks.

All Skills in This Repository (1)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install Empirica?โ–ผ

Run `npx skills add Nubaeon/empirica --all -g -y` in your terminal to install all skills in this suite globally.

What problem does Empirica solve?โ–ผ

It stops AI coding agents from acting before they understand your code, forgetting everything between sessions, and guessing without saying so. It measures confidence, gates edits, and keeps a persistent memory of findings and dead-ends.

Does Empirica work with Claude Code?โ–ผ

Yes. Running the setup command wires hooks, a statusline, and the system prompt into Claude Code automatically, and the skills also load in other SKILL.md-compatible harnesses.

What is the Sentinel gate in Empirica?โ–ผ

It is a checkpoint that blocks code edits until the agent demonstrates enough understanding of the task, using domain-aware confidence thresholds scaled by how critical the work is.

Can I use Empirica without learning its CLI?โ–ผ

Yes. You describe tasks in plain language and the agent operates the 240+ measurement commands itself, while the terminal statusline shows you live confidence and progress.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’