Agent Observability Pattern | Agent Skills Atlas

About this skill

Conventions for observing a multi-step AI SDK agent in a Next.js app — how to model every agent boundary as a closed TypeScript discriminated union, emit it from a single recorder that isolates volatile experimental callback names, stream it as custom data-trace-event parts over the same UI message stream, reduce it in a store, and derive the view through pure selectors with cost computed in one pricing module. Use whenever creating, updating, or porting the src/lib/trace module — trace event union, recorder, pricing, run store, selectors, or the agent run route. Triggers include: 'observe the agent', 'trace events', 'agent timeline', 'token usage', 'tool call timeline', 'stream agent steps', 'cost per run', 'add a trace event', 'instrument the agent'.

Quick start

Copy the install prompt. Use the green button above this README.
Send it to your agent. Paste it into your coding agent’s chat and send.
Review the setup. Your agent checks the skill’s instructions, requirements, and installation scope.

Installation guide →

Review, license & listing details

Review the source before using this skill

This is an imported listing. We haven’t certified its safety, license, or compatibility. Skills may include executable scripts.

Automated security scan: Partial

Only SKILL.md was scanned. Supporting files and scripts were not checked. No safety grade has been assigned.

0 potential findings · Scanned 11/4/2026 · Scanner 0.2.1

Content SHA-256: 8f97785e79c20f433558335e62e737bb82281bc3fd4c645a7a3d881b9b499f42

About scan coverage and limitations

Original listing metadata

Original name

agent-observability-pattern

Original description

Conventions for observing a multi-step AI SDK agent in a Next.js app — how to model every agent boundary as a closed TypeScript discriminated union, emit it from a single recorder that isolates volatile experimental_ callback names, stream it as custom data-trace-event parts over the same UI message stream, reduce it in a store, and derive the view through pure selectors with cost computed in one pricing module. Use whenever creating, updating, or porting the src/lib/trace module — trace event union, recorder, pricing, run store, selectors, or the agent run route. Triggers include: 'observe the agent', 'trace events', 'agent timeline', 'token usage', 'tool call timeline', 'stream agent steps', 'cost per run', 'add a trace event', 'instrument the agent'.

About this listing

Category: Science & research
Reported license: MIT
Listing language (estimated): English
Review status: Not individually reviewed
Agent compatibility: Atlas installation supports Claude Code, Codex, and Cursor. Skill-specific dependencies and features are checked during setup.