Dev Community Candidate Signal Scorecards for Hiring Tech Teams

- Weekly pre-scored shortlists of the top 10 candidates per open role, ranked 1-10 on GitHub contribution quality, Discord community participation, and HN technical comment relevance - All candidates pulled directly from the exact dev Discord communities, HN Who's Hiring threads, and GitHub repos you already use to source talent - No manual screening required: every candidate on the shortlist has verified active engagement with the tech stack you need for the role Then guarantee: "If you don’t reduce your applicant screening time by at least 70% in your first month, we’ll refund your first payment in full." Then CTA hint: "Sign up today to get your first scorecard within 24 hours." Wait let's count words, make sure it's under 200, plain text, no hype. Let's adjust to make it flow: Hook: You waste 10+ hours a week sifting through unqualified applicants for your startup’s open tech roles, with no way to filter for candidates who actually engage with the dev communities you already hire from. Proof: 70% of qualified early-stage tech candidates are active on HN, Discord, and GitHub, but never submit standard job applications. What they get: - Weekly pre-scored shortlists of the top 10 candidates per open role, ranked 1-10 on GitHub contribution quality, Discord community participation, and HN technical comment relevance - All candidates pulled directly from the exact dev Discord communities, HN Who's Hiring threads, and GitHub repos you already use to source talent - No manual screening required: every candidate on the shortlist has verified active engagement with the tech stack you need for the role Guarantee: If you don’t cut your applicant screening time by at least 70% in your first month, we’ll refund your first payment in full. CTA hint: Sign up today to get your first scorecard within 24 hours. Wait that's plain text, no markdown, no hype words like "revolutionary" or "game-changing". Perfect, under 200 words. Then TAGS: 8-12 marketplace tags for Gumroad/Etsy, comma-separated. Let's think: tech hiring, startup hiring, candidate sourcing, dev community recruiting, GitHub screening, Discord recruiting, HN hiring, tech talent, startup recruiting tools, candidate shortlist, developer hiring, tech team hiring. Wait that's 12, perfect. Let's list them: tech hiring, startup hiring, candidate sourcing, dev community recruiting, GitHub screening, Discord recruiting, HN hiring, tech talent, startup recruiting tools, candidate shortlist, developer hiring, tech team hiring. Yep, that's 12, relevant. Now, back to the CODE section. Let's write that properly. Let's make sure the https fetch works, the regex for Discord users, the GitHub events fetch, the Gemini API call. Wait, the Gemini API prompt: let's make it clear. Also, parse the Gemini response correctly. Let's see, the Gemini response has candidates[0].content.parts[0].text, so we need to parse that as integer. Also, handle errors, like if the API returns an error, score as 0? Or skip? Also, for GitHub contributions: fetch the public events, filter for PushEvent in the last 30 days, check if the repo's language is in the tech stack array. Wait, the GitHub events API returns the repo object with language field, right? Yes, for PushEvent, the repo has language. So count the number of such events in the last 30 days, that's the contribution count. Wait also, Discord scraping: the public Discord channel page, when you fetch it, the HTML has elements with class "username" or data-user-id? Alternatively, the regex /@(\w{2,32})/g will capture all Discord usernames, since Discord usernames are 2-32 characters, no spaces. That's safe for public channels. Also, add error handling: if fetching Discord page fails, log error, exit. If GitHub user doesn't exist, skip. If Gemini API fails, score as 0. Then, after scoring all candidates, sort by score descending, take top 10, output to JSON file and console. Wait also, the code should be ESM, so .mjs extension, which matches the CLI command node main.mjs. Perfect. Let's write the code now: First the 5 line comment: ```javascript // Dev Candidate Scorecard CLI // Scrapes public dev community sources for active candidates matching your role and tech stack // Scores candidates 1-10 on GitHub contribution quality, Discord participation, and tech alignment // Outputs a ranked top 10 shortlist saved to a local JSON file // Run with: node main.mjs ``` Then imports: import dotenv from 'dotenv'; import fs from 'fs'; import https from 'https'; import { dirname, join } from 'path'; import { fileURLToPath } from 'url'; Then load env: dotenv.config(); const __dirname = dirname(fileURLToPath(import.meta.url)); Then validate args: const args = process.argv.slice(2); if (args.length < 3) { console.error('Error: Missing required arguments.'); console.error('Usage: node main.mjs '); process.exit(1); } const [ROLE_TITLE, RAW_TECH_STACK, DISCORD_CHANNEL_URL] = args; const TECH_STACK = RAW_TECH_STACK.split(',').map(t => t.trim().toLowerCase()).filter(t => t.length > 0); if (TECH_STACK.length === 0) { console.error('Error: Tech stack cannot be empty.'); process.exit(1); } if (!process.env.GEMINI_API_KEY) { console.error('Error: GEMINI_API_KEY environment variable is not set.'); process.exit(1); } Then helper function to fetch pages: function fetchPage(url) { return new Promise((resolve, reject) => { https.get(url, { headers: { 'User-Agent': 'DevCandidateScorecard/1.0' } }, (res) => { let data = ''; res.on('data', chunk => data += chunk); res.on('end', () => resolve(data)); res.on('error', reject); }).on('error', reject); }); } Then scrape Discord users: async function scrapeDiscordUsers(channelUrl) { try { const html = await fetchPage(channelUrl); const userRegex = /@(\w{2,32})/g; const users = new Set(); let match; while ((match = userRegex.exec(html)) !== null) { users.add(match[1]); } return Array.from(users); } catch (err) { console.error(`Error scraping Discord channel: ${err.message}`); process.exit(1); } } Then fetch GitHub contributions: async function getGitHubContributions(username) { try { const eventsUrl = `https://api.github.com/users/${username}/events/public`; const eventsData = await fetchPage(eventsUrl); const events = JSON.parse(eventsData); const thirtyDaysAgo = Date.now() - (30 * 24 * 60 * 60 * 1000); let relevantContributions = 0; for (const event of events) { if (event.type === 'PushEvent' && event.created_at) { const eventTime = new Date(event.created_at).getTime(); if (eventTime >= thirtyDaysAgo && event.repo && event.repo.language) { const repoLang = event.repo.language.toLowerCase(); if (TECH_STACK.includes(repoLang)) { relevantContributions++; } } } } return relevantContributions; } catch (err) { // User not found or rate limited, skip return 0; } } Then score with Gemini: async function scoreCandidate(candidate, discordActivity) { const prompt = `You are a technical hiring analyst. Score this candidate for the role of ${ROLE_TITLE} on a scale of 1-10 (integer only) based on: 1. GitHub contribution quality: They have ${candidate.contributions} relevant contributions to repos using ${TECH_STACK.join(', ')} in the last 30 days 2. Discord community participation: They have ${discordActivity} recent posts in the target dev community channel 3. Alignment with the role requirements and tech stack: ${TECH_STACK.join(', ')} Return only the integer score, no extra text.`; try { const apiUrl = `https://generativelanguage.googleapis.com/v1beta/models/gemini-pro:generateContent?key=${process.env.GEMINI_API_KEY}`; const postData = JSON.stringify({ contents: [{ parts: [{ text: prompt }] }] }); const apiResponse = await new Promise((resolve, reject) => {