Privacy choices

This site may send the owner a minimal page-visit notice with page, browser, device type, and resolution. If you accept analytics, it can also remember an anonymous visit/session ID and collect richer engagement signals.

Read the privacy notice
AlexPersonal Operating System
Alex OS Online
Back to Skills
LinkedIn
Skill

LLM Observability & Cost Tracing

Used in: alex-os-project (llm_trace.py, /admin/analytics LLM tab)

Short description

Per-call tracing of every LLM request: tokens, model, prompt, answer and USD cost.

Detailed description

Every Claude/OpenAI call in Alex OS is traced to a structured log and rendered in a cost dashboard — no untracked spend.

What I use it for

Knowing exactly what every AI feature costs and how it behaved, surfaced in an admin analytics tab.

Experience context

Knowing exactly what every AI feature costs and how it behaved, surfaced in an admin analytics tab.

Explore with Alex OS

Alex traces every LLM call to a structured log — tokens, model, prompt, response, and USD cost — so no AI spend goes unaccounted.

LLM Observability & Cost Tracing

Per-call logging of every Claude and OpenAI request in Alex OS: captures tokens used, model selected, prompt, response, and dollar cost. Results surface in an admin analytics tab, built in llm_trace.py.

What this covers in practice

Every AI feature cost is visible and attributable — not estimated after the fact. The /admin/analytics LLM tab renders this data directly, making cost and behavior auditable per request.

Skill snapshot

Working knowledge, actively used. Evidence is concrete: a live implementation in alex-os-project with a named file (llm_trace.py) and a dedicated admin view.