Verified:GitHub

instrumentation-planning

Plan observable metrics and spans for AI agent systems using a five-tier framework.

7|1|Updated Dec 26, 2025

Search Tags:#observability #instrumentation #cost-tracking #LLM-tracing #tiered-framework #trace-planning #multi-agent-coordination

Ready to run

npx skills add https://github.com/nexus-labs-automation/agent-observability --skill instrumentation-planning

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: instrumentation-planning
Source: https://github.com/nexus-labs-automation/agent-observability/tree/main/skills/instrumentation-planning
Command: npx skills add https://github.com/nexus-labs-automation/agent-observability --skill instrumentation-planning

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Plan and standardize what to measure in AI agent systems using a tiered, outcome-focused observability framework. It helps teams define metrics, spans, and attribution across foundation, core tracing, context, multi-agent coordination, and evaluation.

Core Features & Use Cases

Tiered implementation framework (Foundation to Evaluation) for scalable observability.
Span naming and attribute conventions to ensure consistent instrumentation across teams.
Sampling guidance, error handling, and production-level guardrails to improve reliability.

Quick Start

Outline an observability plan for a new AI agent system, starting with foundation metrics and progressing through multi-agent coordination and evaluation.

Frequently Asked Questions about instrumentation-planning

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

FAQPage Schema

How do I plan observability instrumentation for multi-agent AI systems?▼

Plan observability instrumentation by applying a tiered framework from foundation to multi-agent coordination, defining standardized spans and attributes to guide consistent measurement across AI systems.

What metrics should I trace for LLM agent observability?▼

LLM tracing should capture foundation metrics, context handling, multi-agent coordination, and evaluation data using standardized span naming conventions to ensure consistent observability across teams.

Does this observability framework support tiered deployment contexts?▼

Yes, the framework explicitly supports tiered deployment contexts, scaling from foundation metrics through core tracing and context up to multi-agent coordination and evaluation tasks.

How do I standardize span naming conventions for AI agent tracing?▼

Standardize span naming by adopting the framework's enforced attribute conventions, which outline specific span and attribute naming rules to maintain consistent instrumentation across teams.

What is the best way to guide instrumentation decisions for AI agents?▼

The best way to guide instrumentation decisions is using an outcome-focused observability plan that defines metrics, spans, and attribution across five tiers from foundation to evaluation.

What guardrails does this framework provide for production observability?▼

The framework provides production-level guardrails including sampling guidance and error handling protocols to improve the reliability of your AI agent observability instrumentation.