Mira is MirrorWeb Insight's AI. Through its chat assistant you can ask Mira, in natural language, to work with your flagging policies and dictionaries. These are the keyword-detection rules that flag communications — not risk policies.

Mira can "List, view, create, or update flagging policies (keyword detection rules)" and "List, view, create, or update flagging dictionaries (phrase/term collections)." Policies contain dictionaries, and dictionaries contain the phrases or terms that flag a message when matched.

One important limit applies to both: update only changes the name. For policies, "update_policy only changes the name; other modifications require the UI." For dictionaries, "update_dictionary only changes the name; phrase changes require the UI." Any deeper edit (changing which dictionaries a policy uses, or editing a dictionary's phrases) must be done in the platform UI.

Select the chat icon in the top navigation bar (tooltip "Toggle chat"). The chat opens with a Policies & Lexicon suggestion, whose prompt is "Help me manage flagging policies and dictionaries".

Ask Mira to create a dictionary, for example "Create a dictionary called Gifts with the phrases 'gift card' and 'voucher'". Mira can create the dictionary with its phrases in a single step. Dictionaries use a match type of "all" or "any" to control how the phrases combine.

Ask Mira to create a policy, for example "Create a global policy named Gifts using the Gifts dictionary". When creating a policy, Mira can set the direction ("both", "inbound", or "outbound"), whether it is global, the platforms it applies to (required when the policy is not global), and which dictionaries it uses.

Ask Mira to rename a policy or dictionary and it will update the name. For anything beyond a rename — changing a policy's dictionaries or editing a dictionary's phrases — Mira directs you to the relevant page to make the change in the UI.