| name | hand-raisers |
|---|---|
| description | Find people publicly asking for what the user sells, this week, across Reddit and forums, qualify each ask, check the competition, and draft a helpful reply. Trigger on /hand-raisers or "find people asking for [X]". |
/hand-raisers: answer people mid-question
Every day, buyers post "anyone got recommendations for...", "which tool for...", "looking for a tool that...". They are publicly asking to be sold to. This skill finds this week's askers in the user's niche, qualifies them, and drafts the reply, while the post is still warm.
Requires the Bright Data MCP server. Nothing else.
Inputs
- What the user sells, in one sentence, and to whom.
- The niche vocabulary: 3-5 words buyers would use in their ask (e.g. "CRM", "cold email", "lead generation"). Derive them from input 1 if the user doesn't specify.
Steps
- Sweep. Bright Data
discover, one call per ask angle, 4-6 calls total. Per call:query: the ask in plain language, e.g.reddit post someone asking for recommendations for a <niche word> toolintent: spell out that you want a genuine question from a buyer, in the last week, not a tutorial, a listicle, or a promotional post.discoverranks on this, so a specific intent is what separates 2 real asks from 20 blog posts.filter_keywords:["reddit.com"]num_results: 20discoverreturns real post URLs, which is what step 3 needs. ⚠️ Passstart_date/end_dateif you like, but do not trust them. Bright Data accepts the range and still returns posts of any age. Freshness is enforced at step 3b, on the real post date, never in the query. ⚠️ Do NOT usesearch_engine/search_engine_batchwithsite:reddit.com "<phrase>" after:<date>. Verified 2026-08-27: the google engine returns encoded/goto?url=CAES...redirect links instead of post URLs, and ignoresafter:. The bing engine returns the Bing interface markup, not results. Both are dead ends for this skill. Optional turbo (KYC accounts only): scrapingreddit.com/r/<sub>/newdirectly requires a Bright Data account with completed KYC (their residential access rule for Reddit). Without it the call returnsResidential Failed (bad_endpoint). If it does, skip it and SAY so; do not retry in a loop. If any Bright Data call times out, retry once before reporting it as unreachable.
- Filter. Results are noisy. Keep only those whose TITLE is a genuine ask in the user's niche (a question, from a buyer, about the category). Drop career advice, hobby posts, vendor self-promotion, and anything off-niche. Expect to keep 5-10 from ~50.
- Read each keeper. Bright Data
web_data_reddit_postswith the post URL. It returns the structured post: author,date_posted,num_comments, full text, and the existing comments with their own dates. 3b. Freshness gate. Drop every post whosedate_postedis more than 7 days old. This is the only place age is decided, because step 1 cannot filter on it. Report the number dropped; a high drop count means the sweep queries were too generic, not that the niche is quiet. - Qualify + competition check. For each survivor: does the ask match what the user sells (yes/partial/no), how fresh is it, and how many real answers does it already have. A 2-day-old ask with 1 comment is gold. A 6-day-old ask with 40 comments is done.
- Draft the reply. Helpful first: answer the actual question, compare 2-3 options honestly, and mention the user's offer as one of them, with the reason it fits THIS ask. No template pitch. The person asked; this is an answer, not an interruption.
Output
A table: ask (title + link) · subreddit · posted when · fit (yes/partial) · answers so far · the drafted reply. Ranked by fit × freshness × low competition. Top 3 first. Below the table, one line: how many results were swept, how many survived the filter, how many were dropped as stale.
Never
- Reply on the user's behalf or post anything. The user copies, edits, posts.
- Publish or store the asker's personal contact info if they left any; point the user to the post instead.
- Force the user's product into an ask it doesn't fit. A bad fit skipped builds more trust than a stretch.
- Present an ask without its real
date_postedfrom step 3. Old asks are dead asks, and step 1 will happily hand you a post from last year.