AWS bedrock-agentcore documentation change
Summary
Multiple formatting and organizational changes including typographical fixes, markdown formatting adjustments, and restructuring of example sections with numbered lists
Security assessment
These changes are primarily documentation formatting improvements: typographical fixes (apostrophes), markdown formatting adjustments (removing backticks, adding spaces), and restructuring example sections with numbered lists. The content change from 'Security' to 'AWS Agent Registry: Centrally catalog and discover AI Resources' appears to be a section title update but doesn't add security documentation. No security vulnerabilities, weaknesses, or security features are mentioned or addressed.
Diff
diff --git a/bedrock-agentcore/latest/devguide/understanding-input-spans.md b/bedrock-agentcore/latest/devguide/understanding-input-spans.md index 21b01d9c1..e21850ede 100644 --- a//bedrock-agentcore/latest/devguide/understanding-input-spans.md +++ b//bedrock-agentcore/latest/devguide/understanding-input-spans.md @@ -22 +22 @@ The `evaluate` API accepts a list of `sessionSpans`, which consists of two types -The evaluation service processes two types of telemetry data to understand your agent's behavior and performance. +The evaluation service processes two types of telemetry data to understand your agent’s behavior and performance. @@ -109 +109 @@ The scope name determines whether the service can process the span. The followin -Span events are associated with spans using `spanId` and `traceId`. The event's scope name is used to determine whether it contains the information required for evaluation. +Span events are associated with spans using `spanId` and `traceId` . The event’s scope name is used to determine whether it contains the information required for evaluation. @@ -144 +144 @@ Span events are associated with spans using `spanId` and `traceId`. The event's -For events with supported scopes, the `body` field follows this schema. **The actual values in the`content` field vary depending on the framework and instrumentation library used.** +For events with supported scopes, the `body` field follows this schema. **The actual values in the content field vary depending on the framework and instrumentation library used.** @@ -179 +179 @@ Below are example spans and their corresponding events for the demo agent create -Strands Agents +###### Example @@ -180,0 +181 @@ Strands Agents +Strands Agents @@ -182 +182,0 @@ Strands Agents -Attribute `"gen_ai.operation.name": "invoke_agent"` is used to identify agent-invocation spans @@ -183,0 +184 @@ Attribute `"gen_ai.operation.name": "invoke_agent"` is used to identify agent-in + 1. Attribute `"gen_ai.operation.name": "invoke_agent"` is used to identify agent-invocation spans @@ -245 +245,0 @@ Attribute `"gen_ai.operation.name": "invoke_agent"` is used to identify agent-in -Langgraph (with opentelemetry-instrumentation) @@ -248 +247,0 @@ Langgraph (with opentelemetry-instrumentation) -Attribute `"traceloop.span.kind": "workflow"` is used to identify agent-invocation spans @@ -249,0 +249,4 @@ Attribute `"traceloop.span.kind": "workflow"` is used to identify agent-invocati +Langgraph (with opentelemetry-instrumentation) + + + 1. Attribute `"traceloop.span.kind": "workflow"` is used to identify agent-invocation spans @@ -298 +300,0 @@ Attribute `"traceloop.span.kind": "workflow"` is used to identify agent-invocati -ExecuteTool span @@ -301 +302,0 @@ ExecuteTool span -Attribute `"traceloop.span.kind": "tool"` is used to identify tool-execution spans @@ -302,0 +304,4 @@ Attribute `"traceloop.span.kind": "tool"` is used to identify tool-execution spa +ExecuteTool span + + + 1. Attribute `"traceloop.span.kind": "tool"` is used to identify tool-execution spans @@ -360 +368 @@ Dataset evaluations -Security +AWS Agent Registry: Centrally catalog and discover AI Resources