Agent Logs Overview

Agent Logs gives you a centralized way to review how your AI agents are behaving across supported Accommador experiences. By bringing execution activity, conversation context, and step-level details into one place, Agent Logs makes it easier to understand what happened, identify issues faster, and improve agent performance over time. This article explains what Agent Logs is, why it matters, and how to use its three-level diagnostic experience to investigate agent activity in Accommador.

What is Agent

Logs? Agent Logs is Accommador’s centralized logging and tracing experience for AI agents. It is designed to give you one unified place to review logs and traces for supported agentic applications in the Accommador ecosystem, helping you understand what is happening behind the scenes during each interaction. This added visibility makes it easier to move beyond guesswork, troubleshoot with confidence, and improve agent behavior with clearer insight into how decisions were made.

As of now, Agent Logs is live for Agent Studio and Voice AI, with support for other AI products coming soon. Agent Logs is built to support a three-level diagnostic experience: A high-level view of agent activityA detailed conversation and execution timeline for a selected interactionA granular step view with deeper technical detail for debuggingKey Benefits of Agent LogsCentralized visibility: Review supported AI agent activity in one place instead of relying only on isolated testing views.

• Faster troubleshooting: Move from a broad activity view into a specific conversation and then into individual execution steps to identify where something went wrong.
• Better context for debugging: Compare the agent’s response with the execution timeline to understand how the final answer was produced.
• Performance awareness: Use execution timing at the step level to spot slow areas that may need attention.
• Improved confidence: Inspect technical details more closely when troubleshooting prompts, logic, variables, or tool behavior.

Global Activity OverviewThe Global Activity Overview gives you a broad view of recent agent activity so you can quickly identify where to focus your attention. This top-level perspective is useful when you want to monitor activity at scale, narrow down a problem, or locate a specific interaction to inspect more closely. The Global Activity Overview serves as the Level 1 experience in Agent Logs.

From here, users can review recent executions and begin narrowing their investigation. This view is useful for: Monitoring recent activity across supported agent experiencesNarrowing down which execution to investigateLocating a specific interaction more quicklyReviewing conversations more efficiently in higher-volume environmentsCommon items and actions in this view may include: Item or ActionDescriptionTimestampShows when the interaction took place so you can review activity in context.

Agent nameIdentifies which agent handled the interaction. Agent IDDisplays the unique ID for the agent tied to the interaction. AI productShows which supported AI product the activity belongs to. ChannelDisplays the channel where the interaction took place.

StatusShows the current result or state of the interaction. Filter by productNarrows results to a specific AI product. Filter by agent nameFocuses on activity for a specific agent. Search by contactHelps find interactions connected to a specific contact.

Sort resultsReorders results to find the most relevant interactions faster. Conversational Context & TimelineSeeing the conversation and the execution history together helps you understand both the user experience and the logic behind it. This view is helpful when you need to answer questions such as why the agent responded a certain way, where a decision changed, or when a tool was involved.

The Conversational Context & Timeline is the Level 2 experience in Agent Logs. After selecting a specific interaction, you can review the conversation alongside an execution timeline. This view helps you: Read the interaction in a familiar conversation formatCompare the agent’s response with the steps that led to itTrace the flow from the original user message through the final responseThe execution timeline is designed to show the ordered path the agent followed during the interaction, giving you a clearer picture of how the conversation progressed behind the scenes.

Task checklist in execution timelines In some Agent Studio runs, the execution timeline includes a task checklist that tracks multi-step work. - Tasks use descriptive labels instead of numeric step indexes. - The UI indicates which task is active and which tasks are completed. - This improves progress clarity when the agent adds tasks or changes the plan mid-run.

Granular Step ExecutionStep-level detail is where troubleshooting becomes much more precise. When a conversation looks incorrect or incomplete, inspecting an individual execution step can help reveal whether the issue came from logic, timing, data handling, or a tool-related action. Depending on the step, users may also be able to review output in different formats, including a raw JSON-style view or a more readable parsed view.

For deeper troubleshooting and sharing, step output can also be copied as JSON. Granular Step Execution is the Level 3 experience in Agent Logs. By selecting a step from the execution timeline, you can review more detailed information about that part of the interaction. This deeper view is useful for examining: The model used for the stepLatency and execution timingTimestamp detailsInput and output for the selected stepPrompt details, where applicableAdditional technical metadata that supports debuggingYou may also see structured task details for multi-step execution (for example, labeled tasks and explicit transitions).

Use these details to confirm task order, completion state, and where progress changed. How To Use Agent LogsAgent Logs does not require a separate installation. The most important part of setup is making sure you have supported agent activity to review and that you know how to move from the high-level view into the detailed execution views. A clear setup process helps you get value from Agent Logs faster and makes troubleshooting more consistent.

Use the steps below to start working with Agent Logs: Open AI Agents >Agent Logs in your Accommador account. Review the activity list and choose the interaction you want to inspect. Use filters like Agent Name, AI Product, Channel, or Status to narrow your results. Open the interaction to view the conversation and execution timeline.

Select a message or step to see more detailed execution information. Use those details to improve your agent. Test again and review updated activity in Agent Logs. Metrics tab (dashboard view)The Metrics tab summarizes Agent Logs data into a customizable dashboard.

Use it to monitor volume, performance, and operational trends across agents, instead of reviewing a single session. Customize the dashboard layoutSelect Edit Layout to change how the dashboard looks: Add or remove widgetsDrag and drop widgets to rearrange the layoutResize widgets to focus on specific charts or tablesSave multiple layouts and switch between them using the layout dropdownWhat Metrics helps you trackMetrics widgets help you monitor: Core KPIs (for example, conversations handled, AI messages, and average messages per conversation)Performance (for example, average response time and top agents)Operational health (for example, top actions and average execution latency)Trends (for example, busiest hours and channel activity over time)For complete widget definitions and dashboard configuration examples, see Agent Logs MetricsFrequently Asked QuestionsQ: What is the main purpose of Agent Logs?

Agent Logs helps you understand how an AI agent handled an interaction by combining conversation context with execution details in a centralized view. Q: Is Agent Logs the same as the Message Execution Timeline in Agent Studio? No. Agent Logs is a centralized logging and tracing experience, while the Message Execution Timeline is part of the Agent Studio testing and debugging experience.

Q: When should I use Agent Logs? Use Agent Logs when you want to investigate how an agent responded, review the path of an interaction, or troubleshoot a specific step in the execution flow. Q: Can Agent Logs help with debugging slow or complex interactions? Yes.

The step-level view is designed to provide more detailed execution insight, including timing and technical context that can help with troubleshooting. Q: Will Agent Logs support other AI products in the future? Yes. Agent Logs is live for Agent Studio at launch, with support for additional AI products coming soon.