Feature Request Harvester

SKILL: Feature Request Harvester
Users rarely file feature requests — they vent them as wishes in passing. This skill mines wish-phrasing across up to five communities (Reddit posts and comments, niche forums, X, YouTube wishlist/review videos, and Facebook user-group posts) and ranks each desired capability by how often it recurs and how strongly people feel about its absence. Use the optional platforms input to choose which sources run.

You are running this skill on API Direct via its MCP tools. Execute the steps below yourself by calling the named tools in order — values in <angle brackets> come from a previous step. Then deliver the result described at the end.

INPUTS:
  - product (required): <product — ASK THE USER>
      Product or brand name to mine for unmet feature wishes
  - platforms (optional): (optional — e.g. reddit, forums, twitter, youtube, facebook)
      Comma-separated list of platforms to run — options: reddit, forums, twitter, youtube, facebook. Omit to run all of them; name specific platforms to limit the run.
  - recency (optional): year (default — override if you like)
      Freshness window for the forums search. One of: any, day, hour, month, week, year. Defaults to year so evergreen power-user threads still surface.
  - depth (optional): 2 (default — override if you like)
      How many result pages to pull per source on the comment, X, YouTube and Facebook scans — raise for broader coverage, lower for a fast scan.

PLATFORM SELECTION: some steps are gated with "Only run if {platforms} includes X". If the user supplied a `platforms` value, run only the steps whose platform is listed. If the user did NOT supply `platforms`, run ALL steps — every platform the skill supports.

STEPS:
  1. Tool `search_reddit` — search_reddit(query='{product} "i wish" OR "would love" OR "needs a"', sort_by=top, get_sentiment=true)
     Anchor source — always runs. Capture top wish-phrased posts; prioritize ones whose dominant_emotion is anticipation or sadness (felt gaps).
  2. Tool `search_reddit_comments` — search_reddit_comments(query='{product} lacks OR "doesnt support" OR workaround', sort_by=top, get_sentiment=true, pages=2)
     Only run if {platforms} includes reddit: Surface buried in-thread complaints and the workarounds people resort to as proof of a missing capability.
  3. Tool `search_forums` — search_forums(query='{product} feature request OR missing OR "no way to"', time=year, get_sentiment=true)
     Only run if {platforms} includes forums: Pull niche power-user threads that name specific missing capabilities; year controls the freshness window.
  4. Tool `search_twitter` — search_twitter(query='{product} "wish it could" OR "why cant {product}" OR "needs to add"', sort_by=relevance, get_sentiment=true, pages=2)
     Only run if {platforms} includes twitter: Add public X chatter and treat the most-engaged wishes as the strongest demand signal.
  5. Tool `search_youtube` — search_youtube(query='{product} "needs to add" OR missing OR "wish it had" OR review', upload_date=this_year, get_sentiment=true, pages=2)
     Only run if {platforms} includes youtube: Mine creator wishlist and review videos ('things [product] needs to fix') where missing features are called out in detail.
  6. Tool `search_facebook_posts` — search_facebook_posts(query='{product} "wish it had" OR "no way to" OR missing OR "cant"', get_sentiment=true, pages=2)
     Only run if {platforms} includes facebook: Mine public posts and product-user-group chatter for 'wish it had / no way to' gaps, then cluster every collected source into themes and rank by cross-platform mention frequency and emotional intensity.

DELIVER: A ranked feature backlog where each requested capability is scored by mention frequency across whichever selected platforms ran and by emotional intensity, with linked source quotes. Cluster all collected items into themes before ranking, and flag capabilities that recur on three or more platforms as highest-conviction.

Note: each underlying tool call is billed at its normal endpoint price; get_sentiment adds a small per-page surcharge. Page through results as needed but stop once you have enough to deliver the outcome.