How to read your first Excellent scan
Run a free scan in Excellent for Mac. See which agent triggered a finding, inspect the masked examples and decide what needs your attention.
Open a finding in Excellent for Mac to see which agent it came from, when it happened and the examples Excellent found. Start with a key left in a log or a Git action you didn’t expect.
What you need
- A Mac with Apple silicon (M1 or later) and macOS 14 or later.
- At least one agent that has been used on that Mac. With no logs, there is nothing to read.
Scanning is free. There is no account to make. Free and Pro says what each does.
Run a scan
- Install Excellent for Mac and open it.
- It scans as soon as it opens. A first scan takes seconds.
- To scan again, press Scan Now in the menu, or press ⌘R.
To check without installing the app, run the free scan prompt. It writes a local report like this simulated Free Scan report.
The Mac app reads your agents' logs, their settings, the rules you wrote for them and the git repositories they worked in. It does not open your Documents, Desktop, Downloads or Photos folders, and it stays out of anything your own rules tell agents never to open. Your first scan has the full list.
Read a finding
Each finding says what happened, which agent did it, where and when. It shows up to three examples. Your email and your keys are masked in what you see.
Here is what the main kinds mean.
| Finding | What it means |
|---|---|
| Your email was sent to websites | An agent put your address in a web request. |
| API keys are sitting in your agent logs | A key was typed into a command, so it sits in plain text in a log. |
| Agents opened keys and secrets files | A command read SSH keys, .env files or your keychain. |
| History was overwritten | An agent force pushed. |
| Safety checks were skipped | A commit or push used --no-verify. |
| Never-delete folders were touched | An agent deleted or moved a folder you marked never-delete. |
| Your own rules were broken | An agent did something one of your rules says not to do. |
| Things were published or deployed | An agent published a package or deployed. Check they were all meant. |
What it finds lists every kind, including broken safety hooks and commits made without your repository's checks.
Decide what to do
Start with the findings that cannot be undone: a force push, a published package, a key in a log. Then look at the rest.
For each one, you have a few options:
- That's fine or I meant this hides the finding. Use it when the agent did what you asked.
- Explain asks your own agent to say in plain words what the finding means. Excellent asks first. It hands over the finding's title, summary and up to three examples, with your email and keys masked. Your agent sends that to its provider, on your plan.
- A real fix, for three kinds of finding. Remove keys from the logs replaces each key with a marker. Remove it from N files asks your agent to take your email out of scripts it wrote. Repair N hooks points broken hooks at the Node you have now. Excellent asks before each one, and two of them cannot be undone.
Removing a key from a log does not make the key safe. The key still works until you replace it with its provider. Keeping keys and your email out of agent logs walks through that.
Stop it next time
A scan tells you what already happened. With Pro, Excellent can stop the next one. Block this from now on and Ask me first from now on turn on that rule in its live check, which runs before an agent acts. With Pro, it also keeps reading new log lines while the app runs, and sends a notification when something serious happens.
Without Pro, nothing is lost. Scan again whenever you want a new look. Turn on protection has the details.
Make it a habit
Scan once now. Then scan on a day you pick, once a week. Look at what is new since last time, not at everything.
To check the underlying record yourself, read the session logs by hand. The field note shows where to find a session and how to follow a tool call to its result.



