shepard-systemshepard-systemOfficialยท6 Agent Skills Included

shepard-obs-stack

Self-hosted observability dashboards for coding assistant usage and cost

Tracks cost, token usage, tool calls, and sessions across Claude Code, Codex, and Gemini CLI. Replaces guesswork about coding assistant spend with nine auto-provisioned Grafana dashboards and real-time alerts. Runs entirely on localhost via Docker, keeping all telemetry data private on your own machine.
npx skills add shepard-system/shepard-obs-stack --all -g -y
Available:

Gives your AI agent the stack architecture, metric naming conventions, and query commands it needs to operate and troubleshoot this observability pipeline correctly.

All Skills in This Repository (6)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install shepard-obs-stack?โ–ผ

Run `npx skills add shepard-system/shepard-obs-stack --all -g -y` in your terminal to install all skills in this suite globally.

How to track AI coding assistant costs?โ–ผ

Start the stack with ./scripts/init.sh, install the hooks, and the Cost dashboard shows spend by provider and model in real time. You can also ask your agent directly with the /obs-cost skill.

Which coding assistants does it support?โ–ผ

It supports Claude Code, Codex, and Gemini CLI through bash hooks plus native OpenTelemetry export. All three feed into the same Prometheus, Loki, and Tempo pipeline.

Does my telemetry data leave my machine?โ–ผ

No. Everything runs locally in Docker containers on localhost, so all metrics, logs, and traces stay on your own machine.

Can I query metrics without opening Grafana?โ–ผ

Yes. The seven included skills let your agent run status checks, cost reports, session history, and free-form PromQL or LogQL queries directly from the terminal.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’