# Observability Setup

> Define local-first logging, metrics, and diagnostic signals for a runtime surface.

- Skill: `ldilov/observability-setup` (Agent Skill)
- Install (CLI): `npx skillmds@latest add ldilov/observability-setup`
- Raw SKILL.md: https://api.skillmd.com/api/skills/ldilov/observability-setup/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: ldilov (https://skillmd.com/u/ldilov)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/ldilov/observability-setup

---


# Observability Setup

## Trigger Signals

- the task needs better local diagnostics or effectiveness tracking

## Inspect First

- runtime entrypoints, existing logs, metrics, and failure reporting surfaces

## Workflow

1. identify the signals worth recording
2. choose low-risk local-first storage
3. define summary and reporting needs
4. recommend the smallest useful setup

## Output Contract

- signal inventory
- storage approach
- reporting approach

## Failure Modes

- the task lacks a clear runtime boundary to instrument

## Escalation

- escalate when observability requirements imply privacy or performance tradeoffs