PreviousNext
AI-assisted engineering/Agent evaluation

Keep only feedback themes backed by enough quotes

Agent buildEasy10 minOpenAI and Anthropic style

Your task

Every quarter, someone spends a week turning user feedback from support, app reviews, sales notes and the forum into themes. An agent clusters the feedback now, and only themes with enough evidence reach the roadmap discussion.

Write select_supported_themes(theme_names, quote_counts, distinct_sources, min_quotes): the check each week's themes pass through.

  • theme_names (list of str): the themes the agent found.
  • quote_counts (list of int): how many verbatim quotes back each theme.
  • distinct_sources (list of int): how many different channels those quotes come from.
  • min_quotes (int): the fewest quotes a theme needs this week.
  • Index i in each list is one theme. Return the names of the supported themes, in rank order.

Rules:

  1. Drop a theme with 0 or fewer quotes, whatever min_quotes is.
  2. Drop a theme with fewer quotes than min_quotes.
  3. Drop a theme whose quotes come from fewer than 2 channels.
  4. Return the rest by quote count, highest first. Equal counts go in ascending name order. A name that appears twice is listed twice.

Examples

select_supported_themes(theme_names: list[str], quote_counts: list[int], distinct_sources: list[int], min_quotes: int) -> list[str]
Example 1
Inputselect_supported_themes(["slow sync", "confusing pricing", "crash on upload"], [12, 5, 3], [3, 2, 1], 4)
Output["slow sync", "confusing pricing"]
Example 2
Inputselect_supported_themes(["export bug"], [9], [1], 3)
Output[]
Example 3
Inputselect_supported_themes([], [], [], 2)
Output[]