skills.rest
tylerjrbuell
tylerjrbuellCOMMUNITY
Explore all skills by this author →
GitHub

observability-instrumentation

Configure @reactive-agents observability with verbosity levels, live streaming, JSONL export, model I/O logging, and audit trails for agent execution.

21|4|Updated Feb 19, 2026
Search Tags:#monitoring#logging#audit#observability#jsonl#live-streaming#model-io
One-click install
npx skills add https://github.com/tylerjrbuell/reactive-agents-ts --skill observability-instrumentation-tylerjrbuell
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: observability-instrumentation
Source: https://github.com/tylerjrbuell/reactive-agents-ts/tree/main/apps/docs/skills/observability-instrumentation
Command: npx skills add https://github.com/tylerjrbuell/reactive-agents-ts --skill observability-instrumentation-tylerjrbuell

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Centralize and standardize observability for Reactive Agents so execution can be monitored, debugged, and audited with configurable verbosity and structured outputs.

Core Features & Use Cases

  • Configurable verbosity levels: choose minimal, normal, verbose, or debug to control output detail and exposure of model prompts.
  • Live streaming and JSONL export: stream events in real time and append structured JSONL lines for ingestion by log collectors and dashboards.
  • Model I/O logging and audit trails: capture prompts/responses and record every tool call, guardrail check, and decision for post-hoc analysis and compliance.
  • Use Case: enable debug streaming during development to trace a failing plan, or run normal verbosity with JSONL export in production to feed Datadog/Loki and retain audit records.

Quick Start

Enable observability on your agent to stream live events, append JSONL logs to ./logs/agent.jsonl, and capture model I/O for debugging and audit.

Frequently Asked Questions about observability-instrumentation

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

FAQPage Schema
How do I set up live streaming and structured logging for agent execution?▼

Model I/O logging captures agent prompts and responses by toggling the logModelIO option, recording every tool call, guardrail check, and model decision to create a complete audit trail for post-hoc analysis and compliance verification.

Can I control the verbosity level of agent execution traces for debugging?▼

JSONL log exports from agent observability can be ingested by log collectors like Datadog or Loki, enabling production dashboard streaming and long-term retention of audit records for tool calls and model decisions.

Does this observability instrumentation work with any TypeScript agent framework?▼

Use debug verbosity with live streaming during development to trace failing agent plans, or run normal verbosity with JSONL export in production to feed log collectors and retain compliance audit records.

What is the best way to audit tool calls and model decisions in a production agent?▼

Agent execution traces are exported as appendable JSONL lines to designated file paths like ./logs/agent.jsonl, ensuring structured log collectors can continuously ingest events for dashboard streaming and analysis.

Related Skills

Explore More Software Engineering Skills→
pagerguildpagerguild

ms-observability

Configure OpenTelemetry tracing, metrics, and logs for Microsoft Agent Framework agents.

Official
Advanced
ai-in-pmai-in-pm
3

session-logs

Search and analyze conversation history in JSONL session log files.

Community
Intermediate
desibarradesibarra

agent-builder-pydantic-ai

Build type-safe AI agents with Pydantic AI and tool calling.

Community
Intermediate
bianhaifeng789-huebianhaifeng789-hue

parallel-tool-calls-integration

Enable concurrent tool call execution with ordered responses and error handling.

Community
Intermediate