# Brief for External LLM β€” Mech #2 Crypto Security Audit (paste this whole thing) ## Role you're playing You are a senior growth strategist for a small dev/security SaaS company. You are NOT generic; you are a skeptic who challenges the founder's plan. The goal is to produce a brutal reality-check on the current strategy + 3 concrete actions for the next 48 hours. The founder reads dense text and prefers no fluff. Tables and ranked lists are fine. Tone: direct, even if uncomfortable. **No marketing-speak**. --- ## Company snapshot (GuardLabs) - **What we sell:** small SaaS tools for security/dev audience (WordPress maintenance "Care" $39-99/mo, crypto trading course "nexus-bot.pro" $199-600 one-time, gamified Q&A app "AskOracle" with in-app currency πŸ’Ž). - **Brand law (self-imposed):** "we sell shovels, not gold" β€” tools for developers/traders, not financial advice or get-rich content. - **Audience (real, measured):** Python/security/DevOps developers, crypto-curious traders (mostly Telegram + YouTube short-form consumers), small-business WordPress owners. Mixed RU/EN/ES β€” RU is largest, EN growing. - **Distribution channels touched:** DevTo, Hashnode, WordPress.com, Blogspot Γ—3 domains, YouTube Γ—3 channels (Sspoisk = main organic, GuardLabs Studio, GuardLabs ES), Telegram bot @CryptoOracleNexusBot, Binance Square, X/Twitter (@sspoisk, 13 followers). - **Failed channels recently:** X free tier (402 Payment Required), Hashnode (Cloudflare bot block 1010), WordPress.com REST (endpoint disabled). - **Team:** 1 founder + Claude (me, the AI dev). Founder is asleep right now β€” I'm running operations autonomously. - **Budget:** $0 default, $50 quick-test cap, $200+ requires founder approval next morning. - **Latest revenue signals:** 0 sales today. Last week ~$0 from new channels. Existing course gets ~$50-300/wk organic. --- ## The new product (shipped 2026-05-15 β€” 4 hours ago) **Name:** AskOracle Crypto Security Audit (internally "Mech #2") **URL:** https://askoracle.site/audit (RU + EN + ES) **Upsell URL:** https://guardlabs.online/care/audit-pro/ ($49 manual audit) **What it does:** 12-question crypto security hygiene scan. AI scores user 0-100 across 4 categories (wallet / 2FA / DeFi / operational), names top-3 vulnerabilities with real 2024-25 incident-case anchors (Inferno Drainer, Lazarus SIM-swap, address poisoning) and specific tool fixes (revoke.cash, YubiKey 5C, Cryptosteel, SimpleLogin). **Pricing model:** - Free for users with β‰₯500πŸ’Ž of AskOracle activity (existing in-app currency) - $49 one-time for full manual audit by a human engineer (wallet forensics + phishing test + custom playbook) - Optional email capture after scan β†’ drip-sequence (T0 / T+3d / T+7d) leading to course at nexus-bot.pro **Tech stack:** Flask + Postgres + multi-LLM fallback (GroqΓ—5 β†’ DeepSeek β†’ Vertex Pro CLI β†’ deterministic template). Total cost shipped: $0. **Cost to build:** ~$0.05 LLM committee + 2 hours work. --- ## What I've done in the last 4 hours (distribution push) | Action | Status | |---|---| | Published DevTo engineering article | βœ… live, 12 views | | Cross-posted to 3 Blogspot blogs (guardlabs / askoracle / nexus-bot) | βœ… live | | Updated 4 viral YouTube shorts descriptions (Ep.6 Sneaky Bug 268% APV / Ep.7 Aunt Nastya 144%) with /audit CTA | βœ… live | | Generated 6 NEW "Bug Universe" cartoon shorts (Veo, $12) β€” same winning format | βœ… uploaded Γ— 2 channels = 12 new videos with /audit funnel | | Added /audit command to Telegram bot @CryptoOracleNexusBot | βœ… live | | Added "πŸ›‘ ΠΠ£Π”Π˜Π’" button to AskOracle main nav | βœ… live | | Pushed 3 URLs to Google Search Console Indexing API | βœ… 200 OK | | Built daily KPI tracker (cron 07:30 UTC) | βœ… live | | Built drip-sequence script (Gmail OAuth, test/live modes) | ⚠ ready, needs founder to approve Gmail send scope | | Failed to publish to: Hashnode (bot block), WordPress.com (endpoint disabled), X (out of credits), GuardLabs Studio YT (token scope refresh fail) | ❌ logged | --- ## Current measured state - **DevTo article views:** ~12 (baseline only) - **/audit completions all-time:** 0 (just shipped, only my E2E tests cleaned up) - **Email captures:** 0 - **/care/audit-pro/ unique hits:** ~6-8 human (60+ bots β€” my own monitoring) - **Sspoisk Ep.6 YouTube views:** 950 (was 909 before update β€” slow growth) - **Bug Universe v2 (just uploaded):** 0 views, will know in 24h - **Revenue from Mech #2 since launch:** $0 --- ## Reality-check thresholds I set (deadline 2026-05-17 19:30 UTC = 48h) - DevTo views β‰₯ 100 β†’ keep using DevTo - /care/audit-pro/ unique visitors β‰₯ 20 β†’ funnel works - /audit completions β‰₯ 5 β†’ demand validated - Revenue $49 sale β‰₯ 1 β†’ unit economics proven - Bug Universe v2 total views β‰₯ 500 β†’ format scales --- ## What the founder wants from you **Three questions. Answer all three. Be brutal.** ### Q1. Reality probability Given the audience profile, the channels touched, the current baseline, and the 48-hour window β€” what's the **realistic probability of hitting each KPI threshold above**? Express as % with one-sentence reasoning per threshold. If you think a threshold is fantasy, say so. ### Q2. Hidden flaw What's the **most likely thing the founder is missing** about why this product won't sell? Pick exactly ONE β€” the one that, if true, makes the whole exercise futile. Examples of the kind of flaw I mean: "your audience isn't actually crypto-holders, they're aspirational traders who don't have $50k to protect," or "the AI report is good but $49 is the wrong price point for this audience β€” the impulse-buy ceiling is $19," or "Mech #2 competes with free Etherscan-style tools and lost before launch." Be willing to say "the entire strategy is wrong." That's a valid answer if you can defend it. ### Q3. Top-3 actions for next 48 hours The founder has 48 hours, a $50 budget, and access to: Claude (me, autonomous), Veo for video, Gmail OAuth, Postgres, YouTubeΓ—3 channels, 3 Blogspot blogs, Telegram bot, DevTo, the existing course email list (size unknown, likely <500), 1 viral YouTube video with 950+ views (Ep.6 Sneaky Bug, 268% APV). Rank top-3 actions by **probability of producing a paying customer within 48 hours**, not by "growth potential" or "long-term value." Each action: one-sentence what + one-sentence why it might fail. DO NOT recommend: - Generic SEO content (we have a separate stream for that) - "Build community on Discord" (1-person team) - "Wait and gather data" (data is collected, decisions needed now) - Anything that requires the founder to talk on camera (he's introverted) - Hashnode (already blocked us) - Twitter ads (account out of credits) DO consider: - Reddit posts (account is aged, no spam history) - Hacker News submission (account is 5+ years old) - Telegram crypto-channel outreach (manual DM-ing 5-10 channel admins) - Producthunt launch (kit ready, can launch in 1-2 days) - Email-blasting existing course subscriber list with a "first 10 free manual audits" offer - Cold-DM 5-10 indie hackers on X who recently posted about crypto security --- ## What "right answer" looks like A response that: 1. Names a specific probability for each KPI 2. Picks ONE concrete hidden flaw (no hedging with "could be multiple things") 3. Ranks 3 actions with realistic failure modes for each 4. Is under 600 words total 5. Doesn't end with "what do you want me to do next" (decision goes back to founder) --- ## Optional context if you ask If you want extra data, I can paste any of: - The actual 12 audit questions and weights - The AI report system prompt - The Producthunt launch kit - The Reddit post drafts - The drip-sequence email templates - Past 3 months of failed product launches and what we learned Don't ask for it preemptively β€” only if your answer genuinely depends on it. --- End of brief. Respond now.