Skip to main content

App Mentions & AI Intent

Mention @frugally in any channel where the app is installed and ask a question in natural language. frugally.app uses AI intent parsing to understand your question and respond with relevant data.


How it works

  1. Type @frugally followed by your question in a Slack channel
  2. frugally.app parses the intent of your message
  3. It queries the relevant data (costs, Executions, Schedules, resources, etc.)
  4. A response is posted in the channel

No slash command syntax required — just ask in plain English.


Supported queries

Cost questions

ExampleWhat it returns
@frugally how much did we spend this month?Total AWS spend for the current month
@frugally what's our EC2 cost this week?EC2 spend for the current week
@frugally are we over budget?Budget status for all active budgets
@frugally what are our top 5 most expensive services?Top services ranked by spend
@frugally compare this month to last monthMonth-over-month spend comparison

Execution questions

ExampleWhat it returns
@frugally what happened last run?Summary of the most recent Execution
@frugally did the dev environment stop last night?Status of the last Execution against dev Targets
@frugally any failed executions today?List of failed Executions from today
@frugally how much have we saved this week?Savings from completed Executions this week

Schedule questions

ExampleWhat it returns
@frugally when does the dev environment stop?Next scheduled stop time for dev Targets
@frugally what schedules are active?List of active Schedules with next run times
@frugally is anything scheduled for tonight?Executions scheduled for the remainder of today

Resource questions

ExampleWhat it returns
@frugally how many EC2 instances are running?Count of running EC2 instances across all Connections
@frugally which resources are not in any target?Resources not covered by any Target
@frugally what's the status of the staging RDS?Current state of RDS resources in staging Targets

Response format

Responses are posted as channel messages and include:

  • Direct answer — The data requested in a readable format
  • Source — Which data source was queried (Cost Explorer, Executions, Schedules, etc.)
  • Link — A link to the full detail in the web dashboard for further exploration
  • Caveats — Any limitations (e.g. "Cost data is up to 24 hours behind")

[SCREENSHOT: slack-ai-mention-response.png -- AI-powered response to a cost question in Slack]


Tips for effective queries

  • Be specific about time — "this month", "last week", "today" help frugally.app return the right data range
  • Name resources — Use Target names or service names to narrow results
  • Ask follow-ups — Mention @frugally again in a thread to ask a follow-up question with context from the previous response
  • Use channels appropriately — Responses are visible to everyone in the channel. For sensitive cost data, use a private channel or DM

What AI intent cannot do

AI intent is read-only. It can answer questions about your data but cannot:

  • Trigger Executions (use /frugally run for that)
  • Create or modify Targets, Schedules, or Budgets
  • Change settings or configurations

If frugally.app cannot understand your query, it responds with a suggestion to rephrase or a link to the relevant dashboard page.

tip

If frugally.app misunderstands your query, try rephrasing with more explicit keywords. For example, instead of "what's happening?" try "what executions ran today?"