A Qt tray-type application that allows for capturing (via user-selectable area or entire window) screenshots, or codeblocks associated with a particular ASHIRT instance. This repo contains just the client. If you are looking for the server you can find it here.
This application allows users to connect to a remote ASHIRT backend and create and submit new evidence. Screenshots are taken using a custom, user-defined key, or alternately by selecting the appropriate action in the tray menu. Codeblocks can be added via an action in the tray. Both can be managed from within the application.
Official releases for Mac and Linux will be provided via the releases tab in GitHub along with source code for users to build themselves if desired.
The official Linux release is delivered as an AppImage. Certain systems may require specialized information. For most distributions, "installation" is easy: simply set the execute flag on the AppImage, then run the application either in a terminal, or via the UI.
| Distribution | Support Status | Notes |
|---|---|---|
| Kali (xfce) | Supported¹ | Status is as of March, 2021 |
| Ubuntu 20.04 (Gnome) | Supported¹ | The operations menu does not update properly |
| Fedora 33 (Gnome) | Partial Support¹² | Key capture may not work properly |
| KDE Neon | Supported¹ | Status as of January 2021 |
| Linux Mint 20.1 (Cinnamon) | Supported¹ | |
| Linux Mint 20.1 (Mate) | Partial Support¹ | An alternate screenshotting tool is needed. e.g. gnome-screenshot |
¹ The application icon does not display properly
² Pure Gnome installations require a taskbar implementation, which is absent on newer versions of gnome. This can be solved by installing gnome extensions. See the list below of compatible Gnome Extensions
Note that, due to Wayland simply operating differently, global hotkeys are not supported on those platforms.
Various gnome extensions work. The below list is a subset. Before using one of these, check that the gnome version they target is the same as the version you are using. Version mismatch could be buggy, or just not work at all
| Extension Name | Notes |
|---|---|
| Tray Icons | |
| Tray Icons: Reloaded | |
| TopIcons Plus |
Current Status: Non-functional
Some OSes/desktops do not support a tray (e.g. ice3 window manager). Currently, in these cases, the application will not work, and simply exit instead. Eventually, a simple CLI will be set up to continue to interact with this application.
On the first launch, the user must first set up an appropriate configuration. When the tray displays, open the tray and select Settings. From here, you will be presented with some options:
| Field | Meaning |
|---|---|
| Evidence Repository | Where evidence is stored. Note that this is a jumping off point. Files are stored in a subdirectory using the operation name |
| Access Key | The (shorter) base-64 key given by the AShirt front end (look for this in Account Settings) |
| Secret Key | The (longer) base-64 key given by the AShirt front end |
| Host Path | The http location to the AShirt server |
| Capture Area Command | The CLI command to take a screenshot of an arbitrary area and save to a file. More on this below |
| [Capture Area Command] Shortcut | The key combination used (at a system level) to trigger the capture area command |
| Capture Window Command | The CLI command to take of a given window, and save to a file |
| [Capture Area Command] Shortcut | The key combination used (at a system level) to trigger the capture window command |
Once the above is configured, save the settings and you can now select an operation. Open the tray, and under Select Operation, choose an operation to start using the application. Note that whenever you change the host path, the list of operations will be updated
This application requires taking screenshots from the command line. The application must:
- Allow for saving the screenshot to a named file.
- Create the file must before the application exits.
- Support a png output format
Theoretically, any application that satisfies this requirement will work. For Mac, the system command to do this is pre-populated, since this is a standard feature. For Linux, there are a number of screenshot commands, and so none are provided. For Windows, a 3rd party application must be used, as there is currently no way to save a screenshot to a named file.
This tool will replace the above filename with %file as noted below:
| OS/DE/App | Capture Window | Capture Area | Notes |
|---|---|---|---|
| Linux/Gnome | gnome-screenshot -w -f %file |
gnome-screenshot -a -f %file |
Capture window captures the focused window, rather than allowing a selection; adding the --delay flag can help mitigate choosing the wrong window |
| Linux/Xfce4 | xfce4-screenshooter -w -s %file |
xfce4-screenshooter -r -s %file |
Capture window captures the focused window, rather than allowing a selection; adding the --delay flag can help mitigate choosing the wrong window |
| Linux/KDE Plasma | spectacle -a -bno %file |
spectacle -r -bno %file |
Capture window captures the focused window, rather than allowing a selection; adding the --delay flag can help mitigate choosing the wrong window |
| MacOS X | screencapture -w %file |
screencapture -s %file |
Note: this application expects a single, basic command. While piping output to another command may work, it is not guaranteed. Likewise, providing multiple commands on the same "line" may work, but is also not guaranteed. Officially, both of these techniques are unsupported. Note 2: Mate-screenshot is unsupported, as it does not appear possible to specify where to write the file without opening up a GUI window
Global shortcut keys can be registered with your computer, depending on the exact operating system. These shortcuts may conflict with shortcuts for a given application, where it is unclear which shortcut will trigger. All this is to say that this feature, while supported, may not work perfectly every time. That said, here is how you configure shortcuts:
Within Settings next to each capture command is a small text box to provide the shortcut. Each shortcut should add in one or more modifier keys (e.g. ctrl) in order to provide less of a chance to interfere with other system/application commands. These modifier keys have reserved names for shortcuts, noted in the below table:
| Key | Name | Alternate Names |
|---|---|---|
| Shift | shift |
shft |
| Control | control |
ctrl |
| Alt | alt |
-- |
| Windows/Meta/MacOS Command key | meta |
win |
To specify a shortcut pattern, simply decide on what set of modifier keys you want, plus a single alphanumeric key (or F- key), and separate these by +.
E.g. Ctrl+Shift+p
To change operations, navigate to Select Operation and choose one of the operations exposed in the list. If the operation you are looking for is not in the list, you can refresh the list by closing and re-opening the tray menu, or check with the operation owner to ensure that you have write access to that operation.
Some users may end up using this application for multiple backends -- either for testing, or due to multiple communities using the same software. Currently this application does not strongly support switching backends, though it is possible. To switch backends, simply enter the new API Key, Secret Key, and Host Path in the settings menu. Note that when switching, it is likely your current selected operation will be removed and you will need to select a new operation.
A particular edgecase that is not supported is when multiple backends have the same name for different operations. In these cases, it is incumbent on the user to be vigilant and ensure that the right data goes to the right backend.
Previous evidence can be reviewed by navigating to View Accumulated Evidence, which will present a screen showing evidence for the current operation. Selecting a row in the evidence list will show:
- A preview of the evidence (Images can be scaled by changing the window size, or my shrinking the description box -- mouse over the divider separating the description from the image)
- The description of the evidence
- Any (active) tags associated with the evidence.
From here you can submit the evidence, if not already submitted. Or, you may delete the file (even if previously submitted -- doing so will remove the file locally, but keep the website copy)
Filtering can be done by specifying items in key:value format. Multiple filters can be added by adding a space between each filter. Keys and values are case insensitive.
| Action | Key | Values | Alias(es) | Notes |
|---|---|---|---|---|
| Show submit errors | err |
t/f, or y/n |
error, fail, failed |
Also works with true/false yes/no |
| Show evidence for operations | op |
operation slug | operation |
Pre-populated with current operation, when reset button is pressed |
| Show evidence taken before a given date | before |
today, yesterday or date in yyyy-MM-dd format, |
to, til, until |
Starts at midnight of the given day |
| Show evidence taken after a given date | after |
today, yesterday or date in yyyy-MM-dd format, |
from |
Start just before midnight of the next given day |
| Show evidence taken on a given date | on |
today, yesterday or date in yyyy-MM-dd format, |
-- | |
| Show evidence that has not been submitted | submitted |
t/f, or y/n |
-- | Also works with true/false, yes/no |
When trying to apply both a "before" date and "after" date filter, the system will adjust the times so that the "before" date is always after the "after" date. Meaning, the timespan must be inclusive. For example, a range of "before March" and "after May" (excluding March and April) is not valid, and will be revised to "After March, Before May"
When applying only one date, the range is unbounded on the other end. That is, dates are implicitly "from the start of time" to "until the end of time"
All data collected from the ASHIRT application can be exported, and then re-imported, into a new ASHIRT instance. Doing so creates a copy on the new system, and the user can pick up where they left off. It is currently recommended that this be used only for moving (rather than copying) data from one computer to the other, when the latter will replace the former. For sharing content, it is recommended that the Web UI be used instead.
To begin an export, open the tray menu, and select Edit > Export. This will open a window where the user can choose a destination, and opt to export only configuration details (specifically, the server connection details), only the accumulated evidence, or both. Finally, press the "Export" button. This will kick off a process that gathers this data, and starts moving it into a central directory for easy migration.
To import content, open the tray and select Edit > Import. This will open a similar dialog to export, but for importing content. Navigate to the export directory, and select the system.json file, and then press the "Import" button. This will kick off a process to bring the exported data into the new system.
Once an import or export has been started, you can close the window. A tray message will display once the action completes. To get progress updates, you can simply reopen the import/export menu. Progress will update once the total number of files is known, and for each file copied.
There are a handful of points to be aware of when importing and exporting.
- You can only export ALL content or NO content Individual asset selection is currently not supported.
- Imports and Exports cannot be cancelled once started
- Creating or editing evidence while importing may be slower The underlying database only allows a single write connection, which means that the import process and main process that allows writing to the local database will need to take turns writing. Depending on your usecase and system, this may or may not delay concurrent work.
- Importing while Exporing (or vice versa) may be confusing Import and export actions are done as a point-in-time action. This means that export will only export what is known to it at the time the "Export" button is pressed. This remains true for import as well, though is less relevant for concurrent actions. As a general peice of guidance, import and export should not be done simultaneously.
- Imports can be re-run, though this is of questionable value Currently there is no way to know if an import has been previously run. Re-importing evidence will cause duplicated files and database records. This does not hamper the system, but is difficult (and manual) to clean up. Please be sure that content has not been previously imported before pressing the import button.
- Exports are unprotected and easily sharable For better or worse, when creating an export file, know that anyone can read or copy this data. Encrypting or decrypting is left as an exercise to the user, should they wish to do so.
- Limitations Currently, the following limitations exist for importing and exporting data:
- "Settings" are not transfered -- specifically, last used tags and operation
- Things that could be operating system dependent are not transfers. This is most of the configuration: hotkey bindings, screenshot commands, and the evidence directory
You should never need to access these files outside of the application, however, for clarity, the following files are generated and maintained by this application:
| File type | Path | Notes |
|---|---|---|
| Screenshots | $eviRepo/$operationSlug/ashirt_screenshot_$randomCharacters.png |
Presently, random (english) characters tacked on to the end of a screenshot, to add uniqueness and prevent overwriting |
| Codeblocks | $eviRepo/$operationSlug/ashirt_codeblock_$randomCharacters.json |
Presently, random (english) characters tacked on to the end of the codeblock filename, to add uniqueness and prevent overwriting |
| Configuration | $userDataDirectory/ashirt/config.json |
Manages connection info / configuration in "settings" menu |
| Local Database | $userDataDirectory/ashirt/evidence.sqlite |
|
| Settings | $userDataDirectory/Unknown Organization/ashirt.conf |
Manages state info -- e.g. last used operation ; Managed by Qt |
The above paths reference some variables. Some of these values change depending on what operating system is being used (or in how it is configured). The exact paths are unknown, but this may help you find these files:
| Path Variable | Notes |
|---|---|
$userDataDirectory |
Where user-specific data files are stored. |
| [For Linux] | On the command line, run echo $XDG_CONFIG_HOME (by default, this is typically the ~/.config directory) |
| [For Mac OSX] | Check /Users/(username)/Application Support/ |
| [For windows] | Check C:\Users\(username)\AppData\Local\ |
$eviRepo |
The Evidence Repository value in the "settings" window |
$operationSlug |
The operation slug for a given operation. This is a unique representation of an operation name |
$randomCharacters |
Six random english characters, case-insensitive (for those operating systems that support this). e.g. fTaNpS |
Interested in contributing? See the developer notes for style guide, organization, etc
All configuration options are managed through the application UI.
Please refer to the contributing.md file for information about how to get involved. We welcome issues, questions, and pull requests.
This project is licensed under the terms of the MIT open source license. Please refer to LICENSE for the full terms.
- Joel Smith
- Alex David
- Joe Rozner
- John Kennedy