[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M76XBADHVCARK5RPSG839E1X":3,"skill-same:01M76XBADHVCARK5RPSG839E1X":45},{"skillNo":4,"name":5,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":20,"frontMatter":21,"refName":22,"commitSha":23,"skillDirectory":24,"forksCount":25,"repositoryHtmlUrl":26,"repositoryHomepageUrl":27,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":29,"updatedAt":30,"install":31,"resources":44},"01M76XBADHVCARK5RPSG839E1X","afrexai-ai-spend-audit",null,"Lord1Egypt\u002Fawesome-skill-forge","Lord1Egypt","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F99746359?v=4","01M7592YDNT0XK9WJ0T2MQVPDN","Mohamed Mounir Ali AL Keshita","01M75GRFNHC8VZNWEB8DKDE433",3,"MIT","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fcommunity\u002Fclawhub\u002Fa\u002Fafrexai-ai-spend-audit\u002FSKILL.md",[],[],"2026-11-12 13:55:43",false,"# AI Spend Audit\n\nAudit your company's AI spending — find waste, measure ROI, and right-size your tool stack.\n\n## When to Use\n- Quarterly AI budget reviews\n- Before renewing AI tool subscriptions\n- When AI spend exceeds 3% of revenue without clear ROI\n- Evaluating build vs buy decisions for AI capabilities\n\n## The Framework\n\n### Step 1: Inventory Every AI Line Item\nMap all AI spending across these categories:\n\n| Category | Examples | Typical Waste |\n|----------|----------|---------------|\n| **Foundation Models** | OpenAI, Anthropic, Google API keys | 40-60% (unused capacity, wrong model tier) |\n| **SaaS with AI** | Salesforce Einstein, HubSpot AI, Notion AI | 30-50% (features enabled but unused) |\n| **Custom Development** | Internal ML teams, fine-tuning, RAG pipelines | 25-45% (duplicate efforts, over-engineering) |\n| **Infrastructure** | GPU instances, vector DBs, embedding compute | 35-55% (over-provisioned, always-on dev instances) |\n| **Data & Training** | Labeling services, training data, synthetic data | 20-40% (one-time costs recurring unnecessarily) |\n\n### Step 2: Score Each Tool (0-100)\n\n**Usage Score (0-30)**\n- 0: Nobody uses it\n- 10: \u003C25% of licensed users active\n- 20: 25-75% active\n- 30: >75% active, daily use\n\n**ROI Score (0-40)**\n- 0: No measurable business impact\n- 10: Saves time but no revenue\u002Fcost link\n- 20: Measurable cost reduction (\u003C2x spend)\n- 30: Clear ROI (2-5x spend)\n- 40: High ROI (>5x spend)\n\n**Replaceability Score (0-30)**\n- 0: Commodity (10+ alternatives at lower cost)\n- 10: Some alternatives exist\n- 20: Few alternatives, moderate switching cost\n- 30: Irreplaceable, deep integration\n\n**Action Thresholds:**\n- Score 0-30: **CUT** — cancel immediately\n- Score 31-50: **REVIEW** — renegotiate or find alternative\n- Score 51-70: **OPTIMIZE** — right-size tier\u002Fusage\n- Score 71-100: **KEEP** — monitor quarterly\n\n### Step 3: Model Cost Optimization\n\nFor every API-based AI tool, check:\n\n1. **Model Selection**: Are you using GPT-4 where GPT-3.5 suffices? Claude Opus where Sonnet works?\n   - Rule: Use the cheapest model that meets quality threshold\n   - Test: Run 100 production queries through cheaper model, measure quality delta\n\n2. **Caching**: Are you re-processing identical or similar queries?\n   - Semantic cache can cut 20-40% of API calls\n   - Exact-match cache catches another 5-15%\n\n3. **Batch vs Real-time**: Which requests actually need sub-second response?\n   - Batch processing is 50% cheaper on most providers\n   - Queue non-urgent requests for batch windows\n\n4. **Token Optimization**:\n   - Trim system prompts (every token costs money at scale)\n   - Use structured output to reduce response tokens\n   - Implement max_tokens limits per use case\n\n### Step 4: Vendor Consolidation\n\nMap overlapping capabilities:\n\n```\nCurrent State → Target State\n─────────────────────────────────────────\nChatGPT Teams + Claude Pro + Gemini → Pick ONE primary + ONE backup\nJasper + Copy.ai + ChatGPT for content → Single content tool\n3 different vector databases → Consolidate to 1\nInternal embeddings + OpenAI embeddings → Standardize on one\n```\n\n**Consolidation savings**: Typically 25-40% of total AI spend.\n\n### Step 5: Build the Audit Report\n\n```\nAI SPEND AUDIT — [Company Name] — [Quarter\u002FYear]\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\nTotal AI Spend: $___\u002Fmonth ($___\u002Fyear)\nAI Spend as % Revenue: ___%\nIndustry Benchmark: 2-5% (early adopter) \u002F 0.5-2% (mainstream)\n\nWASTE IDENTIFIED\n├── Unused licenses: $___\u002Fmonth\n├── Over-provisioned infra: $___\u002Fmonth\n├── Model tier downgrades: $___\u002Fmonth\n├── Vendor consolidation: $___\u002Fmonth\n└── TOTAL RECOVERABLE: $___\u002Fmonth ($___\u002Fyear)\n\nACTIONS\n┌─ CUT (Score 0-30): [list tools]\n├─ REVIEW (Score 31-50): [list tools]\n├─ OPTIMIZE (Score 51-70): [list tools]\n└─ KEEP (Score 71-100): [list tools]\n\n90-DAY PLAN\nWeek 1-2: Cancel CUT items, begin REVIEW negotiations\nWeek 3-4: Implement model downgrades and caching\nWeek 5-8: Vendor consolidation migration\nWeek 9-12: Measure savings, establish ongoing monitoring\n```\n\n## Company Size Benchmarks (2026)\n\n| Company Size | Typical AI Spend | Typical Waste | Recoverable |\n|-------------|-----------------|---------------|-------------|\n| 10-25 employees | $2K-$8K\u002Fmo | 35-50% | $700-$4K\u002Fmo |\n| 25-50 employees | $8K-$25K\u002Fmo | 30-45% | $2.4K-$11K\u002Fmo |\n| 50-200 employees | $25K-$80K\u002Fmo | 25-40% | $6K-$32K\u002Fmo |\n| 200-500 employees | $80K-$300K\u002Fmo | 20-35% | $16K-$105K\u002Fmo |\n| 500+ employees | $300K-$1M+\u002Fmo | 15-30% | $45K-$300K\u002Fmo |\n\n## Red Flags\n\n- AI spend growing faster than revenue (unsustainable)\n- More than 3 overlapping tools in same category\n- No usage tracking on AI SaaS licenses\n- GPU instances running 24\u002F7 for dev\u002Ftest workloads\n- Paying for enterprise tiers with startup-level usage\n- No A\u002FB testing between model tiers\n- \"Innovation budget\" with no success metrics\n\n## Industry Adjustments\n\n- **SaaS\u002FTech**: Higher AI spend acceptable (5-8%) if it's in the product\n- **Professional Services**: Focus on billable hour impact — $1 AI spend should save $5+ in labor\n- **Manufacturing**: AI spend should tie to defect reduction or throughput gains\n- **Healthcare**: Compliance costs inflate spend 20-30% — factor in before judging waste\n- **Financial Services**: Model risk management adds 15-25% overhead — legitimate cost\n- **Ecommerce**: Measure AI spend per order — should decrease as volume scales\n\n---\n\n*Built by [AfrexAI](https:\u002F\u002Fafrexai-cto.github.io\u002Fcontext-packs\u002F) — AI operations context packs for business teams. Run the [AI Revenue Calculator](https:\u002F\u002Fafrexai-cto.github.io\u002Fai-revenue-calculator\u002F) to find your biggest automation opportunities.*\n",{},"master","4aa004da461d820f4ea5499e58eb8d196cf99016","community\u002Fclawhub\u002Fa\u002Fafrexai-ai-spend-audit",0,"https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge","https:\u002F\u002Fskill-explorer-rho.vercel.app","https:\u002F\u002Fgithub.com\u002FLord1Egypt","Unbreakable. Unstoppable. Unmatched⚡.\r\n Speaker of machine tongues 🤖  | World-builder of decentralized dreams 🌐 | Defender of the digital realm🛡️.","2026-06-10 08:21:01",{"source":7,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add Lord1Egypt\u002Fawesome-skill-forge --skill afrexai-ai-spend-audit",{"tool":37,"command":38},"bunx","bunx skills add Lord1Egypt\u002Fawesome-skill-forge --skill afrexai-ai-spend-audit",{"tool":40,"command":41},"pnpm","pnpm dlx skills add Lord1Egypt\u002Fawesome-skill-forge --skill afrexai-ai-spend-audit","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fcommunity\u002Fclawhub\u002Fa\u002Fafrexai-ai-spend-audit\u002FSKILL.md",true,[],[46,54,61,68,75,82],{"skillNo":47,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":50,"categories":51,"keywords":52,"publishedAt":53},"01M785WKJB99G3STR1NF9NRTJW","youtube-content","YouTube transcripts to summaries, threads, blogs.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fyoutube-content\u002FSKILL.md",[],[],"2026-11-15 14:19:17",{"skillNo":55,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":53},"01M785WEFHB0G2S9YA6EPR3H0E","xurl","X\u002FTwitter via xurl CLI: post, search, DM, media, v2 API.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fxurl\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":53},"01M785W98EDN4EFJXJPMQEHCRC","touchdesigner-mcp","Control a running TouchDesigner instance via twozero MCP — create operators, set parameters, wire connections, execute Python, build real-time visuals. 36 native tools.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Ftouchdesigner-mcp\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":53},"01M785W5S9FP3QBH2V191E025P","test-driven-development","TDD: enforce RED-GREEN-REFACTOR, tests before code.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Ftest-driven-development\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":53},"01M785W2AS08NSSH9YATCV6FJB","teams-meeting-pipeline","Operate the Teams meeting summary pipeline via Hermes CLI — summarize meetings, inspect pipeline status, replay jobs, manage Microsoft Graph subscriptions.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fteams-meeting-pipeline\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":53},"01M785VXKW4T34BTP1SCHW1R7F","systematic-debugging","4-phase root cause debugging: understand bugs before fixing.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fsystematic-debugging\u002FSKILL.md",[],[]]