Yesterday's Work, Summarized From Git by an Agent

Preparation for a daily standup meeting often consumes more mental energy than the meeting itself. When you are deep in a complex feature branch, remembering every specific change made over the previous eight hours is challenging. Utilizing an agent to create a yesterday's work summarized from git report allows you to bridge the gap between commit history and human readable status updates. This approach ensures accuracy while letting you remain focused on solving technical problems instead of manual documentation.

The Problem With Manual Status Updates

Software engineering requires intense focus. Interrupting a flow state to scrub through bash history, Jira tickets, or git logs to prepare for a ten minute meeting is inefficient. The friction of context switching often leads to vague updates like "worked on the API" or "fixed some bugs."

These generic updates provide little value to the team. A better approach involves using automation to parse terminal history and version control metadata. By delegating the retrieval of context to an agent, you ensure that your report includes specific file changes, refactors, and logic updates that might otherwise be forgotten.

Constructing a Git Based Reporting Pipeline

A reliable summary starts with raw data. Git provides various ways to view recent activity, but the git log command alone can be noisy. To get a useful summary, you need to filter for specific timeframes and format the output so an AI agent can interpret it effectively.

Useful Git Commands for Summarization

Before asking an agent to synthesize information, you should understand the commands that provide the best signals.

• git log --since="24 hours ago" --oneline: Provides a concise list of commit messages from the last day.
• git log --since="yesterday" --patch: Shows the actual code changes, which is vital if your commit messages are brief or non-descriptive.
• git diff @{yesterday}..HEAD: Compares the state of the repository from yesterday to the current moment.
• git status: Identifies uncommitted work that is currently in progress.

When these outputs are fed into a large language model or a specialized coding agent, the resulting summary becomes much more detailed. The agent can categorize changes by feature, identify potential regressions, and even suggest what the next logical steps should be.

Using AI Agents for Contextual Reporting

Modern developer tools allow you to run agents directly within your workspace. These agents have access to your file system and terminal, making them ideal for generating reports.

In a local-first environment, your code and history stay on your machine. This is crucial for privacy and security when using AI tools for internal work. Tools like Claude Code, Codex CLI, or OpenCode can run locally and examine your recent branches without uploading your entire repository to a remote server.

Integrating Deska Into Your Workflow

Deska provides an infinite canvas where you can organize these reporting tasks alongside your primary editor and terminal. Because Deska is a desktop application for Mac, Windows, and Linux, it handles the execution of these agents in a secure environment.

Running Agents Side by Side

One of the strengths of Deska is the ability to run multiple coding agents in parallel panels. You can have a terminal running a build process in one corner and an agent like Claude Code in another.

To generate a summary, you can use Ask Deska, the built-in assistant. By asking the assistant to check your recent git history and summarize the changes found in specific terminals, you get a formatted report ready for your team. This process uses the local context of your machine, ensuring that your data and storage remain private.

Monitoring From Your Phone

If you are away from your desk when a standup begins, the mobile app allows you to check in. You can remotely trigger a summary command or view the output of an agent session through a secure relay. Since the devices pair directly, you do not need to expose any ports to the internet to get your status update on the go.

Best Practices for Agent Driven Summaries

To get the most out of an automated summary, follow these guidelines.

1. Descriptive Commits: Even the best agent struggles with a commit message that just says "fix." Use clear messages to give the agent a better starting point.
2. Contextual Range: Limit the agent to the last 24 to 48 hours to avoid overwhelming the context window with irrelevant history.
3. Review Before Sending: AI can occasionally misinterpret the importance of a specific change. Always read the summary before pasting it into a shared channel.
4. Use Notebooks: Store your favorite reporting prompts in notes within your workspace for quick access setiap morning.

Comparison of Local vs Cloud Summarization

Choosing where the summarization happens affects both speed and security.

• Cloud Based Summarizers: These often require a browser extension or a GitHub integration. They work well for teams but require sending code metadata to a third party.
• Local Agents: These run on your hardware. They have access to local files that haven't been pushed yet. This is the preferred method for privacy conscious developers.
• Hybrid Workspaces: Tools that offer remote access tools while keeping primary logic local offer the best of both worlds.

Deska supports a BYOK model for its lifetime tier. This means you can use your own API keys for providers like Anthropic or OpenAI, giving you full control over the costs and the models used for your standup reporting.

FAQ

How can I get a git summary for today?

You can run git log --since="00:00:00" --pretty=format:"%s" to see all commit subjects from today. If you use a tool like Deska, you can ask the agent to specifically look at today's work and format it into a bulleted list for a specific Slack channel or meeting.

Is it safe to give an AI agent access to my git history?

If you are using a local-first tool, the agent runs against your local files. While the prompt and the git log text might be sent to an LLM provider if you are using a cloud API, the files themselves stay on your machine. For maximum security, use local inference models that stay entirely within your network.

Can agents summarize work from multiple branches?

Yes, agents can be instructed to compare your current branch with a main branch or to look across all branches modified in the last 24 hours. This is especially helpful if you are context switching between a bug fix and a feature build and need a unified report for both.

Try Deska for Your Next Standup

Automating your reporting workflow keeps you in the code and out of the documentation churn. By using a flexible workspace that supports multiple AI agents, you can customize exactly how your work is summarized and presented.

You can download the free desktop app for Mac, Windows, or Linux today. Start by setting up a new canvas, opening a terminal panel, and asking the assistant to help you prepare for your next meeting.