For the complete documentation index, see llms.txt. This page is also available as Markdown.
New

Logs & Data

Trace model calls, API requests, and database operations.

Trace the platform activity behind each outcome. Logs & Data records model calls, platform API traffic, external requests, and database operations.

Trace requests

Inspect incoming platform requests and outgoing calls to external services.

Investigate execution detail

Review model consumption and database operations behind platform behavior.

At a glance

Area
Use it to
Start here

LLM Logs

Trace model usage, tokens, cost, and latency

Request Logs

Inspect incoming and outgoing API traffic

DB Query Logs

Audit database operations

All four screens support workspace and date-range filters and search. Every row expands to a full detail view.

LLM Logs

LLM Logs records every LLM call completion with its model, target, tokens, and cost. Open Logs & Data → LLM Logs.

The LLM Logs screen, listing model calls with tokens, cost, latency, and status
LLM Logs, tracing every completed model call.

You can narrow by workspace and model, and search by model or agent. Each row is one completed model call:

Field
What it shows

Model

The model that ran (for example, gpt-4.1-mini).

For

The target the call was made for.

Tokens

Tokens consumed by the call.

Cost

The credit cost of the call.

Latency

How long the call took, in milliseconds.

When

The completion time.

Status

The call outcome (for example, Success).

Select a row to see the call's full detail.

Incoming Request Logs

Incoming Request Logs records the requests received by the platform API. Open Logs & Data → Incoming Request Logs.

The Incoming Request Logs screen, listing requests with method, URL, and timestamp columns
Incoming Request Logs, the default view.
Incoming Request Logs with the HTTP method filter open, showing GET, POST, PUT, PATCH, and DELETE options
Filtering Incoming Request Logs by HTTP method.
Incoming Request Logs with the module filter open, showing Agent, App, Skill, Integrations, Marketplace, and Platform options
Filtering Incoming Request Logs by module.

You can narrow by workspace, by one or more modules — Agent, App, Skill, Integrations, Marketplace, or Platform — by one or more HTTP methods — GET, POST, PUT, PATCH, or DELETE — and by date range, and search the log. Each row shows the request's Method, its URL, and When it arrived. Select a row for the full request detail.

Outgoing Request Logs

Outgoing Request Logs records the requests the platform made to external services. Open Logs & Data → Outgoing Request Logs.

Outgoing Request Logs with the module filter open, showing status and response time columns
Outgoing Request Logs, filtered by module.

The same ways to narrow apply — workspace, module, method, and date range. Each row adds a status and timing:

Field
What it shows

Method

The request method (for example, GET, DELETE, or CUSTOM for calls the platform can't map to a standard verb).

Status

The HTTP response status.

URL

The destination the platform called.

Response Time

How long the external call took, in milliseconds.

When

When the call was made.

DB Query Logs

DB Query Logs records the database operations the platform executed. Open Logs & Data → DB Query Logs.

The DB Query Logs screen, listing operations with module, operation, user, and timestamp columns
DB Query Logs, the default view.
DB Query Logs with the module filter open, showing Agent, App, Skill, Integrations, Marketplace, and Platform options
Filtering DB Query Logs by module.

You can narrow by workspace and by module — Agent, App, Skill, Integrations, Marketplace, or Platform — and search the log. Each row is one operation:

Field
What it shows

Module

The area the operation belongs to (for example, Agents, Guardrails, Knowledge, MCP, Accounts).

Operation

The operation, in schema.table.action form (for example, agents.agent.update).

User

The actor that triggered it.

Timestamp

When it ran.

Control CenterObservabilityCredits & TokensControls & Audit

Last updated