← ClaudeAtlas

painpointlisted

Check whether anyone actually has the problem behind an idea. Searches six public sources for people describing it right now, pulls out whoever already shipped it (including repos with thousands of stars), and drafts an opener you edit and send yourself. Says COULD NOT SEARCH rather than reporting a false zero. No API key.
VGokulsai/painpoint-finder · ★ 1 · AI & Automation · score 75
Install: claude install-skill VGokulsai/painpoint-finder
# painpoint Someone has an idea. Before they spend months on it, three questions matter: **is the problem real, has somebody already solved it, and can I talk to one of these people today?** This skill answers all three from public data, with no API key and no account. ## When to use this Use it the moment someone describes something they want to build — before writing any code for it. That is the whole point: a search costs a minute, and building the wrong thing costs months. Also use it when someone asks whether an idea is "taken", who the competitors are, or where to find users to talk to. **Do not** use it to decide for them. It reports evidence. It does not know whether an idea is good, and saying so is part of using it correctly. ## Setup (first run only) The tool is a small Python program. The base directory of this skill is given to you when the skill loads — use that path literally. Check whether the searcher is already there, and clone it if not: ```bash cd "<the skill base directory you were given>" [ -d painpoint-finder ] || git clone --depth 1 https://github.com/VGokulsai/painpoint-finder.git python -c "import requests" 2>/dev/null || pip install requests ``` Do **not** try to derive the directory with `$0` or `dirname` inside this file. Shell variables written here are substituted before the block ever reaches a shell — `$0` came out as the first character of the user's argument the first time this ran, which would have pointed the clone at a directory