observability

Implement PII-safe telemetry and local diagnostic logging for desktop applications.

5|Updated Apr 1, 2026
One-click install
npx skills add https://github.com/TechyMT/claude-code-superpowers --skill observability-techymt
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: observability
Source: https://github.com/TechyMT/claude-code-superpowers/tree/main/skills/observability
Command: npx skills add https://github.com/TechyMT/claude-code-superpowers --skill observability-techymt

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Observability establishes a clear separation between local diagnostic logs and telemetry to prevent accidental exposure of user data while providing actionable analytics and startup performance insights. It prevents PII leakage to analytics, ensures consistent token cost accounting, and adds lightweight profiling so engineers can pinpoint import and initialization regressions.

Core Features & Use Cases

  • Local structured debugging using a local-only debug logger for rich context that never goes to telemetry.
  • PII-safe telemetry events via typed analytics metadata and event naming conventions for cross-boundary operations.
  • Startup profiling and token cost tracking with checkpoint timers and a shared cost accumulator for UI and internal policies.
  • Use Case: Instrument a new tool to emit a PII-safe completion event, write detailed error traces only to local logs, and add profile checkpoints around initialization phases.

Quick Start

Replace console.log calls with the local debug logger, add PII-safe analytics events for operation outcomes, and insert profileCheckpoint markers to measure startup timing.

Frequently Asked Questions about observability

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I implement PII-safe telemetry in my local agent application?

To implement PII-safe telemetry, enforce typed analytics metadata and strict event naming conventions. This separates local diagnostic logs from telemetry pipelines, preventing accidental user data exposure during operation outcomes.

What is the best way to track token usage costs during tool calls?

The best way to track token usage costs is accumulating data in a shared cost tracker. This enforces consistent token cost accounting across I/O operations and tool calls, supporting UI display and internal policy enforcement.

How do I profile startup performance and initialization regressions?

Profile startup performance by inserting profileCheckpoint markers around initialization phases. This adds lightweight profiling to record duration, helping engineers pinpoint import and initialization regressions in desktop applications.

How do I separate local debug logs from analytics events?

Separate local debug logs from analytics by replacing console.log calls with a local-only debug logger. Write detailed error traces exclusively to local logs while emitting PII-safe analytics events for operation outcomes.

Does this observability pattern work for desktop applications?

Yes, this observability pattern works for desktop applications. It instruments I/O operations, startup phases, and tool calls specifically for desktop or local-agent environments, ensuring local structured debugging without telemetry exposure.

Why should I enforce typed analytics metadata for telemetry events?

Enforce typed analytics metadata to ensure PII-safe telemetry events. Typed metadata maintains a clear separation between local diagnostic logs and telemetry, preventing accidental user data exposure while providing actionable analytics.