Claude Reflect is a beta feature that helps you review how you have been using Claude, including the topics you discuss, the tasks you delegate, your most active periods, and patterns in how you work with AI.
The useful part is not simply seeing a dashboard. It is using that reflection to ask a more important question: Is AI supporting your thinking, or quietly replacing parts of it that you still want to practise?
Current feature status: Anthropic introduced Claude Reflect on July 9, 2026. It is rolling out in beta for eligible Free, Pro, and Max accounts using Claude on the web or Claude Desktop. Memory must be enabled. It is not currently available for Team or Enterprise plans, and the Reflect page is not yet available in the Claude mobile app.
What Claude Reflect Can Tell You About Your AI Use
Claude Reflect is a personal recap of your recent Claude activity. Instead of listing every conversation, it turns your chat history into a higher-level view of the subjects you worked on, when you tended to use Claude, and how you collaborated with it.
Your recap is designed to reflect your patterns back without judging them. You can then decide whether those patterns support your goals, whether you are delegating the right tasks, and whether any part of your AI routine needs to change.
Your Activity Patterns
The recap can show your most active day, peak hour, total conversations, and a daily activity chart for the selected period.
Your Common Topics
Claude groups the themes you spent time on and may show a proportional breakdown with a short description of each topic.
Your Working Habits
The report can identify patterns such as rewriting drafts in your own voice, providing context clearly, or deciding on a strategy before delegating execution.
Your AI Fluency
Observations are organized around four skills: Delegation, Description, Discernment, and Diligence.
How to Find Claude Reflect
You do not need to install a separate tool. The monthly recap is built into the Claude settings for eligible accounts.
- Open Claude on the web or Claude Desktop. The Reflect page is not currently available directly through Claude Mobile.
- Open Settings and select Reflect. The default recap loads activity from the previous month.
- Choose the period you want to review. Depending on the available menu, you can review this month so far, the past month, the past three months, the past six months, or the past year.
- Refresh the report when needed. Use the refresh control to regenerate the recap for the selected period.
- Check your memory setting if Reflect is missing. Open Settings → Capabilities and confirm that memory from chat history is enabled.
Why you may not see Claude Reflect yet
The feature is rolling out gradually. It may also be missing because memory is off, you are using a Team or Enterprise account, or you are looking for the Reflect page inside the mobile app.
Mobile conversations can still contribute to an eligible recap even though the recap page itself is not currently available on mobile.
What the Claude Reflect Dashboard Shows
A current Claude recap is divided into several sections. The exact wording may vary because Claude generates parts of the report from your own chat history.
Opening Summary
The opening section gives you a short headline and description of your recent use. It should refer to the actual subjects and types of work found in your conversations rather than giving you generic praise.
Activity Overview
The dashboard currently includes your most active day, peak hour, total number of conversations, and an activity chart showing how your conversations were distributed across the selected period.
Topic Breakdown
Claude can group the subjects and themes you worked on, describe each one briefly, and estimate the proportion of your activity connected with each topic.
AI Fluency Observations
The recap may point out how you tend to delegate, describe tasks, evaluate answers, and take responsibility for AI-assisted work. It can also suggest a practical next step, such as creating a Claude Project or improving the context you provide for recurring work.
Claude Reflect does not currently show exact total time spent
Anthropic has said that a view showing how much time you have spent using Claude is planned, but it is not currently part of the live feature. Do not confuse total conversations, peak hour, or daily activity patterns with a verified total number of hours.
Claude-generated summaries and percentages can occasionally be inaccurate. Treat the report as a useful reflection, not an audited record. If a topic percentage or description looks wrong, use the feedback option provided on the page.
The Four AI Fluency Skills in Claude Reflect
Claude Reflect uses Anthropic’s 4D AI Fluency Framework to describe how you work with AI. These four skills are more useful than simply measuring how often you open an AI tool.
Delegation
Choosing what to hand to AI, what to keep human-led, and what outcome you are actually trying to achieve.
Description
Giving Claude enough context, goals, constraints, examples, and instructions to produce a useful result.
Discernment
Evaluating whether an AI response is accurate, relevant, complete, trustworthy, and appropriate for the situation.
Diligence
Taking responsibility for how you use AI, how you represent its contribution, and what happens after you act on its output.
Seven Questions to Ask About Your Claude AI Habits
A dashboard is only useful when it leads to a better decision. Use these seven questions to turn Claude Reflect from an interesting recap into a practical monthly review.
What tasks do you delegate most?
Look at your common topics and recurring tasks. Are you using Claude mainly for research, organization, drafting, planning, coding, learning, decision support, or repetitive work?
Delegation is not automatically a problem. The important question is whether you intentionally chose to delegate the task or gradually stopped doing it yourself without noticing.
Identify the three tasks you delegate most and decide whether each one should remain AI-assisted, become human-led again, or use a combination of both.
Are you checking important AI-generated answers?
AI output can sound confident even when a detail is incorrect, outdated, incomplete, or misunderstood. Verification matters most when an answer affects money, work, education, safety, health, legal rights, customer information, or another person.
Spot-check dates, numbers, names, quotations, sources, calculations, and important recommendations before relying on them.
Does the final work still sound like you?
AI can help you create a first draft, but your experiences, judgment, tone, examples, and perspective are what make the final result genuinely yours.
If every email, report, caption, assignment, or article begins and ends with an untouched AI response, your own communication style may gradually become less visible.
Rewrite the opening, add one personal observation, remove generic phrases, and read the final version aloud before using it.
Which skills do you still practise without AI?
Skills become stronger through use. Writing, research, problem-solving, memory, creative thinking, planning, and decision-making can weaken when you consistently skip the parts that require effort.
You do not need to avoid AI. You need to preserve opportunities to think, learn, and create independently.
Pick one important skill and practise it independently at least once each week before asking AI to review or improve your work.
Are sensitive conversations appearing in your patterns?
Claude Reflect may refer to personal or sensitive subjects at a high level. Sensitive or distress-related topics should not lead the recap or appear as itemized counts and percentages, but this does not mean you should share unnecessary confidential information.
Avoid entering passwords, payment details, private client data, confidential business files, identification documents, or personal information that is not required for the task.
Would quiet hours or break reminders help?
Claude includes optional Time and focus controls. A break reminder can nudge you after a chosen amount of daily use, while quiet hours remind you that you set a particular period aside for something else.
These controls are not hard locks. You can stop, snooze, dismiss, or continue using Claude.
Set boundaries around sleep, family time, focused work, prayer, study, exercise, or any activity that benefits from fewer digital interruptions.
What one habit will you change next month?
A long list of restrictions is difficult to maintain. One clear adjustment is more useful than a dramatic plan you abandon after a few days.
Choose one change, such as verifying sources, drafting the first paragraph yourself, keeping one evening AI-free, or limiting connected-app access.
What Should AI Help With—and What Should Stay Human-Led?
There is no universal list that works for every person. A low-risk repetitive task can be a sensible use of AI, while a personal, ethical, confidential, or high-consequence decision usually needs more human control.
| AI can reasonably assist with | Keep human judgment in charge |
|---|---|
| Organizing rough notes into a clear structure | Deciding what you truly believe or want to communicate |
| Creating a first-draft outline | Approving the final message, claim, or recommendation |
| Summarizing non-sensitive information | Checking whether the summary is accurate and fair |
| Brainstorming several possible approaches | Choosing the approach that fits your values and circumstances |
| Formatting repetitive text or data | Protecting private information and controlling permissions |
| Breaking a large task into smaller steps | Accepting responsibility for the outcome |
| Explaining a difficult concept in simpler language | Learning, practising, and demonstrating your own understanding |
When an AI tool can access email, files, calendars, browsers, or connected services, the decision is not only about productivity. It is also about permissions and data exposure. Review the AI Agent Safety Checklist before giving an assistant broad access to your apps.
Claude Reflect Privacy: What Is Included and Excluded?
Claude Reflect is built from the same recent chat history used by Claude memory. If memory is turned off, the recap is hidden. Anthropic says the recap is generated when you visit the Reflect page rather than continuously creating a visible report in the background.
Conversations currently excluded from the recap
- Incognito chats are skipped.
- Conversations using health integrations, such as Apple Health or Health Connect, are skipped.
- Claude Cowork activity is currently skipped.
- Claude Code activity is currently skipped.
Connected services and files
Claude Reflect does not pull the raw email or original file from a connected service into the recap. However, text Claude created about that content can influence the recap.
For example, if Claude summarizes your inbox, the original emails are not placed inside the recap, but the summary Claude produced may be represented as part of your activity.
Sensitive subjects
Personal or difficult topics may still be reflected at a high level. Anthropic says sensitive or distress-related topics should not lead the recap or be itemized with counts and percentages.
This treatment does not remove the need for normal privacy caution. Only share information that is necessary, review your account and memory settings, and understand the permissions of any service you connect.
How to Set Claude Quiet Hours and Break Reminders
Quiet hours and break reminders are separate from the recap, although the Reflect page can link you to them. You can use these controls without opening Claude Reflect.
Set a break reminder
- Open Settings → Time and focus. Use Claude on the web or Claude Desktop to change the setting.
- Choose your daily time threshold. Select how much time with Claude should pass before you receive a reminder.
- Respond to the reminder. You can stop, snooze the reminder, dismiss it, or continue working.
Set quiet hours
- Open Settings → Time and focus. Find the Quiet hours section.
- Select the relevant days. Choose the days on which you want the reminder to apply.
- Choose a start and end time. Claude will remind you when you open it during that period.
Remember: quiet hours and break reminders create light friction, not a hard block. They are personal boundaries you can dismiss or continue past, and they are separate from your plan’s usage limits.
A Simple Monthly AI Habit Review
You do not need a complicated tracking system. Once each month, open your recap and answer five practical questions.
- Which tasks did AI genuinely make easier or faster?
- Which AI-generated claims did I accept without checking?
- Did my final work still include my own voice and perspective?
- Which important skill did I continue to practise independently?
- What single boundary or habit will I change next month?
You can use the free AI Daily Task Planner to organize your day and separate tasks that benefit from AI assistance from tasks you want to complete yourself.
Useful monthly reflection prompt
Copy this prompt after reviewing your Claude Reflect recap:
Help me review my AI habits without judging or praising me. Here are the patterns I noticed in my Claude Reflect recap: [PASTE OR SUMMARIZE YOUR PATTERNS] Ask me: 1. Which tasks I delegate most 2. Which answers I may not be checking carefully enough 3. Which skills I still want to practise myself 4. Whether my final work still sounds like me 5. Whether any privacy or connected-app risks need attention Then help me choose one realistic habit to improve before next month. Do not make the decision for me. Help me compare the options and keep final judgment under my control.
Signs You May Be Delegating Too Much to AI
Frequent AI use does not automatically mean unhealthy or irresponsible use. The better question is whether the way you use it is reducing your understanding, independence, privacy, or accountability.
You may need to adjust your routine when:
- You regularly submit or publish AI output without reading it carefully.
- You cannot explain the reasoning behind work produced in your name.
- You ask AI to make personal decisions before identifying your own priorities.
- You use AI to avoid practising a skill you genuinely want or need to learn.
- Your messages and writing no longer sound like your natural voice.
- You share confidential information simply because copying everything feels faster.
- You stop checking sources because the answer sounds polished.
- You allow AI to send, delete, purchase, publish, or change information without appropriate review.
The solution is not necessarily to stop using AI. It may be enough to move from automatic use to intentional use: define the task, limit access, check the output, add your own thinking, and keep responsibility for the final action.
Not Sure Whether a Task Should Stay Human-Led?
Compare the benefits, risks, consequences, and level of human judgment required before you delegate it.
Use the Free Decision Helper Explore More Everyday AI GuidesOfficial Sources and Further Reading
Claude features can change while they are in beta. Check the official documentation if you are reading this guide later or if your account displays different options.
Anthropic Feature Announcement
Official introduction to Claude Reflect, its purpose, AI fluency framework, privacy approach, and planned additions.
Read Anthropic’s Claude Reflect announcementClaude Monthly Recap Help
Current access requirements, recap sections, eligible plans, exclusions, and troubleshooting information.
View the official monthly recap guideTime and Focus Controls
Official instructions for setting, changing, or removing quiet hours and break reminders.
View quiet hours and break reminder instructionsAnthropic Privacy Policy
Broader information about personal data, connected services, user choices, retention, and account privacy.
Review Anthropic’s privacy policyFrequently Asked Questions About Claude Reflect
What is Claude Reflect?
Claude Reflect is a beta monthly recap that summarizes how you use Claude. It can show common topics, activity patterns, your most active day, peak hour, total conversations, and observations about how you delegate tasks, describe goals, evaluate answers, and use AI-generated work.
Where can I find Claude Reflect?
Open Claude on the web or Claude Desktop, go to Settings, and select Reflect. Memory from chat history must be enabled under Settings → Capabilities.
Is Claude Reflect free?
Claude Reflect is currently available in beta to eligible Free, Pro, and Max users. It is not currently available on Team or Enterprise plans.
Why can’t I see Reflect in my Claude settings?
Confirm that you are using Claude on the web or Claude Desktop, that your account is on a Free, Pro, or Max plan, and that memory is enabled. The feature is also rolling out gradually, so some eligible users may receive it later.
Does Claude Reflect show how many hours I have used Claude?
Not currently. The live recap shows activity information such as your peak hour, most active day, total conversations, and daily conversation patterns. Anthropic has said that a view of total time spent using Claude is planned for a future update.
Does Claude Reflect include incognito chats?
No. Anthropic states that incognito chats are excluded from the recap. Conversations involving health integrations, Claude Cowork, and Claude Code are also currently excluded.
Does Claude Reflect read files from connected services?
The recap does not incorporate the original raw email or file from a connected service. However, content Claude produced about that material, such as an inbox summary, may be represented in the recap.
Can sensitive conversations appear in Claude Reflect?
Sensitive or personal subjects may appear at a high level. Anthropic says distress-related topics should not lead the recap or be itemized with counts and percentages.
Are Claude quiet hours a hard lock?
No. Quiet hours and break reminders are optional personal controls. You can stop, snooze, dismiss the reminder, or continue using Claude.
Does opening Claude Reflect use my plan allowance?
Viewing the recap does not count toward your usage limit. Starting a new Claude conversation from a suggestion on the recap page counts as a normal conversation.
Use AI Deliberately—Not Automatically
The goal is not to use AI as little as possible. The goal is to understand where it genuinely helps, where its output needs checking, and where your own thinking still deserves space.
Let AI assist with speed, structure, and repetitive work. Keep your judgment, responsibility, creativity, learning, and personal voice active.
Fact-checked against official Anthropic information on July 13, 2026. Claude Reflect is a beta feature, so availability and functionality may change.
