菜单
Calculate unit economics for your AI product. Revenue per action, API cost per action, gross margin, and free tier damage. Run when model pricing changes or you're rethinking your credit pricing.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | economics |
| description | Calculate unit economics for your AI product. Revenue per action, API cost per action, gross margin, and free tier damage. Run when model pricing changes or you're rethinking your credit pricing. |
| category | ai |
| tags | ["economics","pricing","margins","ai","cost"] |
| author | tushaarmehtaa |
Calculate the current unit economics for your AI product.
Search for credit/pricing constants:
grep -rn "CREDIT\|credit_cost\|COST\|FREE_CREDITS\|signup_bonus\|price\|PRICE" --include="*.py" --include="*.ts" -i
Find:
Search the web for current pricing for every model you use. Common ones to look up:
site:openai.com pricingsite:anthropic.com pricingsite:ai.google.dev pricingGet input $/1M tokens, output $/1M tokens, and image generation cost if applicable.
Make reasonable assumptions based on your prompts (or read the actual prompt files):
If you can find the actual prompts in the codebase, read them and estimate more precisely.
For each user-facing action:
API cost = (input_tokens / 1M × input_price) + (output_tokens / 1M × output_price)
Revenue = credits_charged × price_per_credit
Margin = (revenue - api_cost) / revenue × 100
UNIT ECONOMICS — [your product]
════════════════════════════════════════════════════════════════
USER PRICING
[price] = [credits] credits → $X.XX per credit
Signup bonus: [N] free credits
ACTION ECONOMICS
────────────────────────────────────────────────────────────────
Action Credits Revenue API Cost Margin
────────────────────────────────────────────────────────────────
[action name] [N] $X.XX $X.XXX XX%
[action name] [N] $X.XX $X.XXX XX%
────────────────────────────────────────────────────────────────
FREE TIER DAMAGE (N signup credits)
Conservative user: costs you ~$X.XX
Typical user: costs you ~$X.XX
Heavy user: costs you ~$X.XX
PAYBACK POINT
One paying user covers ~X non-converting free users
════════════════════════════════════════════════════════════════
Show what margins look like at ±20% credit price. Helps decide whether to change pricing without rebuilding the whole spreadsheet.
Wire PostHog analytics, Sentry error tracking, and a health endpoint into any web project. Reads your stack, installs the right packages, and instruments your app so you stop flying blind in production.
Full Answer Engine Optimization audit — make your project discoverable by AI search engines (ChatGPT, Perplexity, Claude, Google AI Overviews), not just Google. Use when the user wants to improve their site's visibility in AI-powered search, add structured data, create llms.txt, optimize for AI citations, or audit AEO readiness. Triggers on requests like "AEO audit", "AI search optimization", "make my site visible to AI", "structured data", "schema markup", "llms.txt", "answer engine", "AI discoverability", or any mention of being found by AI search engines or chatbots.
Audit or generate all UI microcopy for an app — empty states, error messages, loading states, onboarding flows, tooltips, button labels, confirmation dialogs, and success messages. Use when the user wants to write UI copy, fix empty states, improve error messages, audit UX writing, make their app feel polished, or replace placeholder text. Triggers on requests like "write app copy", "fix my empty states", "error messages", "UI copy", "microcopy", "UX writing", "my app feels robotic", "placeholder text", "button labels", "onboarding copy", "toast messages", or any mention of in-app text, UI writing, or product copy.
Generate a weekly changelog card for Twitter/X. Reads git log, distills the work into 6 punchy items, fills an HTML template, and opens it in the browser ready to screenshot.
Write cold emails that actually get responses. Covers investors, employers, mentors, partners — anyone you don't know yet. Strict 2-4 sentence formula with real examples that got replies from Evan Spiegel, Mark Cuban, and Elon Musk. Triggers on requests like "cold email", "write an email to...", "reach out to...", "email this investor", "email this founder", "outreach email", "networking email", or any request to contact someone you don't have a relationship with yet.
Generate complete multi-touch cold outreach sequences — not just one email, but a full 3-5 email sequence with timing, escalation logic, and a different angle per touch. Use when the user needs a sales sequence, drip campaign, follow-up series, or any cold outreach with more than one touchpoint. Triggers on requests like "outreach sequence", "email sequence", "follow-up series", "drip campaign", "sales cadence", "multi-touch outreach", or when someone needs more than a single cold email.