Whatever message this page gives is out now! Go check it out!

AI Services monitoring tab and widget reference

Last update:
May 18, 2026
Find column definitions, widget descriptions, and field details for every tab in AI Services monitoring within the ColdFusion Performance Monitoring Toolset.

Agents tab

Slowest agent requests grid shows individual slow agent executions.
ColumnDescription
Template pathFile or template that handled the request.
OperationType of work (for example chat or ingest).
Duration (ms)How long the execution took.
Line noSource line associated with the measurement.
StatusOutcome. Rendered as a color-coded badge (green for success, red for error).
TraceClickable View Trace action. Opens the Trace Details modal overlay.
Most frequent agents grid shows the most invoked template and operation combinations.
ColumnDescription
Template pathFile or template.
OperationType of work.
Avg duration (ms)Average execution time.
CountNumber of executions.
URL performance table shows per-URL latency and hit counts on the Agents tab.
ColumnDescription
URLRequest URL.
ApplicationApplication name.
Average response time (ms)Mean latency for that URL.
HitsNumber of requests.

LLMs tab

Slowest LLM calls grid lists individual high-latency LLM invocations.
ColumnDescription
ModelModel identifier (for example claude-sonnet-3-5).
ProviderProvider name (for example ANTHROPIC, OPEN_AI).
Duration (ms)Wall time for that call. Sorted descending by default.
TokensTokens attributed to that call (prompt + completion).
TraceOpens Trace Details for that call.
Most frequent models grid shows which models run most often.
ColumnDescription
ModelModel identifier.
ProviderProvider name.
Avg duration (ms)Average latency.
CountNumber of invocations.
LLMs tab distribution pies summarize traffic, tokens, and status.
WidgetDescription
Token usage by providerHow tokens split across providers.
Provider distributionHow requests split across providers.
Model distributionTraffic by model ID.
Status distributionOutcomes for LLM calls.

Vector Stores tab

The Vector Stores tab has two on-screen sections: Vector Queries and Embeddings.
Slowest vector queries grid lists the individual slowest vector query executions.
ColumnDescription
StoreVector store name (for example InMemory).
OperationQuery operation name.
Duration (ms)Query execution time.
Template pathSource template.
Line noSource line.
StatusOutcome.
TraceOpens Trace Details.

RAG tab

Completed Ingest Operations grid shows finished ingestion runs.
ColumnDescription
Template pathSource template.
OperationIngest operation name.
Duration (ms)Total operation time.
DocumentsNumber of documents processed.
Line noSource line number.
StatusOutcome.
View detailsOpens the Ingest Operation Details slide-out panel.
In-Progress Ingest Operations grid shows operations that have not finished.
ColumnDescription
Template pathSource template.
OperationIngest operation name.
Start timeWhen the operation started.
Elapsed (ms)Time the run has been running so far (not total duration).
DocumentsNumber of documents.
Line noSource line.
StatusCurrent status.

Trace Details fields

Summary bar fields appear below the tabs in the Trace Details modal.
FieldDescription
DurationTotal time for the trace.
StatusOverall outcome. Rendered as a color-coded badge.
TemplateFull path to the source file.
LineLine number tied to the trace entry.
Timeline span TypeLabels identify the kind of work each span represents.
TypeLabelDescription
AgentTop-level agent request span.
LLMTime in the language model call.
EmbeddingEmbedding work.
VectorVector store access.
ToolTool call execution.
GuardrailGuardrail evaluation.
RAG PhaseA RAG pipeline phase span.
MCP ServerMCP server-side operation.
Basic Info tab fields show read-only summary data for the trace.
FieldDescription
OperationHigh-level operation name.
DurationTotal duration, matching the summary bar.
Total tokensToken usage when the platform records it.
Retriever configuredWhether a retriever was set up (Yes / No).
Has toolsWhether tool calling was involved (Yes / No).
StatusSuccess or failure.
Memory TypeShown only when a memory type is configured.
ErrorShown only when status is an error.
All Spans tab columns list every span in the trace.
ColumnDescription
TypeSpan category (for example aiservice, llm).
Service / ProviderLogical service or model identifier.
OperationOperation name for that span.
NameSpan-specific name (for example, the tool name for Tool spans).
Duration (ms)Time spent in that span. Sortable.
StatusPer-span outcome.

MCP Clients tab

Slowest MCP Client Operations grid lists the highest-latency MCP client invocations.
ColumnDescription
Server NameMCP server identifier.
OperationOperation type.
NameTool name.
Duration (ms)Call duration.
StatusOutcome.
TraceOpens Trace Details.
Most Frequent Operations By Server grid shows which tools run most often per server.
ColumnDescription
Server NameMCP server identifier.
OperationOperation type.
NameTool name.
Avg Duration (ms)Average call duration.
FrequencyNumber of invocations.

MCP Servers tab

Slowest MCP Server Operations grid lists the slowest server-handled invocations.
ColumnDescription
Server NameMCP server identifier.
OperationOperation type.
NameTool name.
Duration (ms)Execution time.
StatusOutcome.
Note: Trace links are not available from the MCP Servers tab. To trace a call, switch to the MCP Clients tab and locate the same operation there.
Most Frequent Operations By Server grid shows which tools run most often from the server perspective.
ColumnDescription
Server NameMCP server identifier.
OperationOperation type.
NameTool name.
Avg Duration (ms)Average execution time.
FrequencyNumber of invocations.

Share this page

Was this page helpful?
We're glad. Tell us how this page helped.
We're sorry. Can you tell us what didn't work for you?
Thank you for your feedback. Your response will help improve this page.

On this page