Contributing to Wikipedia involves some invisible friction: finding content that needs improving, understanding what guidelines apply, and gathering the right context can all be effortful parts of the process. Suggestion Mode is a VisualEditor feature designed to reduce this friction while also helping people to publish constructive changes that they might not otherwise. It does so by proactively suggesting various types of edits one can consider making to improve content in an article.

Some of the suggestions apply broadly across all wikis, like detecting pasted text, inappropriately located external links, and images without captions. But what makes a constructive change can vary in practice beyond these shared cases. The patterns that signal a potential problem in an edit are rooted in the language, culture, and local editorial history of each project, and no single global set of suggestions can fully account for that diversity.
Rather than trying to solve this centrally, the Editing team is working to give communities the tools to define what edit suggestions would best identify these signals for their wiki.
TextMatch: a tool for custom suggestions
TextMatch is a type of Suggestion that searches for predetermined queries and, when a match is found, displays the regular Edit Check dialog along with the corresponding message to the editor, title, and actions for that query. Here are two screenshots that show a locally defined type of Suggestion, and the result in an example article.


Implementations of TextMatch – called ‘match rules’ – can use plain text or regex (regular expression) queries along with any combination of configuration options, such as: ignoring quoted content; ignoring or including specific sections, article categories, and templates; and defining the upper and lower bounds of edit counts of the users that the suggestion should be shown to.
TextMatch is entirely locally-customizable, so communities can use TextMatch to create custom suggestions that are tailored to their own policies and writing conventions. Its capabilities are perhaps best shown in the range of suggestions that several communities have already built with it.
What communities are building
Russian, Hebrew, Czech, Italian, and Ukrainian wikis have written a combined 80+ custom match rules. The types of issues that these rules have been written to detect include: common typos, cliches, first-person pronouns, second-person verb usage, relative time keywords (such as “currently”), non-neutral keywords or phrases, and many more.
Power users have emerged at Russian Wikipedia in particular, where two volunteers have made a combined 200+ edits to MediaWiki:Editcheck-config.json, which is the configuration file for TextMatch and all other types of suggestions.
The issues that Russian Wikipedia’s rules detect include:
- Outdated country names
- Incorrect island prepositions
- A paragraph that is too long
- Unreliable sources
- And many, many common errors related to football
A number of these Suggestion types have utilized a combination of TextMatch’s regex capabilities and the different configuration options for suggestions. For example, the “old-country-names” rule suggests replacing outdated country names with their modern equivalents, but excludes articles about years before 1991:

The Editing team is encouraged to see how much local knowledge is being reflected in these custom suggestions, and we invite other communities to try it out and share what you build.
Join in
- Read the TextMatch documentation
- View all existing suggestions on your wiki at
Special:EditChecks(E.g. English Wikipedia’s Special:EditChecks), and see the current configuration on your wiki in theMediaWiki:Editcheck-config.jsonlinked at the top. - Share feedback
- And if you have an idea that you believe is too complicated for TextMatch, add it to the list of edit check ideas!
Thanks to all contributors who have provided feedback on TextMatch – and Suggestion Mode as a whole – and are helping us continue to improve the experience for communities. A special thanks to users Well very well and Mitte27, two volunteers at Russian wiki who have been doing extensive work with TextMatch. The depth of their usage and the thoughtful feedback they’ve provided have been invaluable in helping us refine the feature.
Taking a step back, the Editing Team sees TextMatch as an early milestone in its broader effort…
- …to equip volunteers with the technical capabilities they need to reduce the toil involved with contributing constructively to Wikipedia.
- …around empowering experienced volunteers to offer newcomers the clarity and guidance they need to edit constructively in the moments they need it most…while they’re editing!
Can you help us translate this article?
In order for this article to reach as many people as possible we would like your help. Can you translate this article to get the message out?
Start translation