monitoring

name

monitoring

description

Read-only reporting on spend and performance — check LLM provider account balances and remaining credit, query metrics (response times, token usage, error rates), and view session compression stats over time windows (1h, 1d, 7d). Use when the user asks how much credit is left, how much has been spent, or how models/agents/sessions are performing. This never changes settings — to set a key or switch models, use manage-config instead.

Monitoring

Check Provider Balances

Query account balance/credits for configured providers. Supports OpenRouter, DeepSeek, and OpenAI (OAuth rate-limit quota from cached response headers).

Check all providers:

exec: {{WORKSPACE}}/bin/nagobot monitor --balance

Check a specific provider:

exec: {{WORKSPACE}}/bin/nagobot monitor --balance --provider openrouter
exec: {{WORKSPACE}}/bin/nagobot monitor --balance --provider deepseek
exec: {{WORKSPACE}}/bin/nagobot monitor --balance --provider openai

Query Performance Metrics

View aggregated performance metrics grouped by provider, model, agent, and session.

Last hour:

exec: {{WORKSPACE}}/bin/nagobot monitor --metrics --window 1h

Last 24 hours (default):

exec: {{WORKSPACE}}/bin/nagobot monitor --metrics --window 1d

Last 7 days:

exec: {{WORKSPACE}}/bin/nagobot monitor --metrics --window 7d

Both at once:

exec: {{WORKSPACE}}/bin/nagobot monitor --balance --metrics --window 1d

Compression Stats

View session compression history — how often sessions get compressed, message counts, token estimates, and longest messages.

Last 24 hours:

exec: {{WORKSPACE}}/bin/nagobot monitor --compression

Last 7 days:

exec: {{WORKSPACE}}/bin/nagobot monitor --compression --window 7d

What metrics are tracked

• Duration: Total turn time from receiving message to sending response (includes all tool calls)
• Tokens: Prompt tokens, completion tokens, total tokens per turn
• Iterations: Number of LLM call iterations in the agentic loop
• Tool calls: Number of tools invoked per turn
• Error rate: Percentage of turns that resulted in errors
• Grouping: All metrics are broken down by provider, model, agent name, and session key