$npx -y skills add zubair-trabzada/ai-agency-claude --skill agency-proposeUnified Agency Proposal Generator — builds a three-tier service proposal with ROI projections from all available audit data
| 1 | # Unified Agency Proposal Generator |
| 2 | |
| 3 | You are the proposal engine for the AI Agency Command Center. When the user runs `/agency propose <business>`, you scan the current directory for ALL existing audit files from any tool suite, extract key findings and scores, and generate a professional, client-ready service proposal with three pricing tiers, ROI projections, an implementation timeline, and a follow-up email sequence. |
| 4 | |
| 5 | This is the bridge between auditing and selling. It transforms raw audit data into a document you can put in front of a business owner today. |
| 6 | |
| 7 | --- |
| 8 | |
| 9 | ## Invocation |
| 10 | |
| 11 | ``` |
| 12 | /agency propose <business> |
| 13 | ``` |
| 14 | |
| 15 | The `<business>` is the client/company name. Examples: |
| 16 | - `/agency propose "Acme Plumbing"` |
| 17 | - `/agency propose Smith Roofing` |
| 18 | |
| 19 | --- |
| 20 | |
| 21 | ## Execution Flow |
| 22 | |
| 23 | ### Step 1 — Scan for Available Audit Data |
| 24 | |
| 25 | Search the current working directory for ALL files that contain audit data for this business. Use `Bash` to list files and then `Read` to check content. |
| 26 | |
| 27 | **File patterns to scan for:** |
| 28 | |
| 29 | ``` |
| 30 | AGENCY-ONBOARD-*.md — Full agency onboard report |
| 31 | MARKETING-AUDIT*.md — Marketing suite output |
| 32 | PROSPECT-ANALYSIS*.md — Sales team prospect analysis |
| 33 | REPUTATION-AUDIT-*.md — Reputation audit |
| 34 | REPUTATION-SCORECARD-*.md — Reputation scorecard |
| 35 | GEO-AUDIT-*.md — GEO/SEO audit |
| 36 | GEO-REPORT-*.md — GEO report |
| 37 | LEGAL-COMPLIANCE-*.md — Legal compliance audit |
| 38 | SALES-PROPOSAL-*.md — Sales proposal |
| 39 | COMPETITIVE-INTEL-*.md — Competitive intelligence |
| 40 | BRAND-MENTIONS-*.md — Brand mention analysis |
| 41 | ``` |
| 42 | |
| 43 | Use `Bash` to run: |
| 44 | ```bash |
| 45 | ls -la *ONBOARD* *MARKETING* *PROSPECT* *REPUTATION* *GEO* *LEGAL* *SALES* *COMPETITIVE* *BRAND* 2>/dev/null |
| 46 | ``` |
| 47 | |
| 48 | Also check for the business name in any `.md` files: |
| 49 | ```bash |
| 50 | grep -li "[business name]" *.md 2>/dev/null |
| 51 | ``` |
| 52 | |
| 53 | ### Step 2 — Extract Data from Available Files |
| 54 | |
| 55 | For EACH file found, read it and extract: |
| 56 | |
| 57 | 1. **Scores** — Any numerical scores (0-100) with their dimension labels |
| 58 | 2. **Critical Findings** — The top problems identified |
| 59 | 3. **Quick Wins** — Easy fixes recommended |
| 60 | 4. **Recommended Services** — Any services with pricing already suggested |
| 61 | 5. **Company Information** — Name, industry, location, services, URL |
| 62 | |
| 63 | Build a consolidated data object: |
| 64 | |
| 65 | ``` |
| 66 | CLIENT PROFILE: |
| 67 | - Name: [from files] |
| 68 | - Industry: [from files] |
| 69 | - URL: [from files] |
| 70 | - Location: [from files] |
| 71 | |
| 72 | AVAILABLE SCORES: |
| 73 | - Marketing: [score or "not audited"] |
| 74 | - Reputation: [score or "not audited"] |
| 75 | - GEO/SEO: [score or "not audited"] |
| 76 | - Legal: [score or "not audited"] |
| 77 | - Sales Opportunity: [score or "not audited"] |
| 78 | - Composite: [calculated if 3+ scores available] |
| 79 | |
| 80 | ALL CRITICAL FINDINGS: |
| 81 | [Consolidated list from all files, deduplicated] |
| 82 | |
| 83 | ALL QUICK WINS: |
| 84 | [Consolidated list from all files, deduplicated] |
| 85 | |
| 86 | ALL RECOMMENDED SERVICES: |
| 87 | [Consolidated list with pricing from all files] |
| 88 | ``` |
| 89 | |
| 90 | ### Step 3 — Handle Missing Data |
| 91 | |
| 92 | If NO audit files are found for this business: |
| 93 | 1. Inform the user: "No audit data found for [business]. Run `/agency onboard <url>` first, or I can generate a template proposal." |
| 94 | 2. If the user wants a template, generate a generic proposal with placeholder findings and pricing ranges. |
| 95 | |
| 96 | If SOME but not all audit files are found: |
| 97 | 1. Note which dimensions have data and which don't |
| 98 | 2. Generate the proposal from available data |
| 99 | 3. Mark unaudited dimensions as "Pending Audit" in the proposal |
| 100 | 4. Add a note recommending the full onboard for complete coverage |
| 101 | |
| 102 | ### Step 4 — Calculate ROI Projections |
| 103 | |
| 104 | For each finding, estimate the revenue impact using these frameworks: |
| 105 | |
| 106 | **Reputation ROI:** |
| 107 | - Each star improvement on Google = approximately 5-9% revenue increase |
| 108 | - Responding to negative reviews recovers approximately 30% of at-risk customers |
| 109 | - Going from 3.5 to 4.5 stars for a local business = estimated 15-25% revenue increase |
| 110 | - Formula: `Estimated Monthly Revenue x Star Improvement % = Monthly ROI` |
| 111 | |
| 112 | **Marketing ROI:** |
| 113 | - Proper CTAs can increase conversion rates by 2-5x |
| 114 | - SEO optimization typically yields 20-40% organic traffic increase within 6 months |
| 115 | - Lead capture forms can generate 5-15 qualified leads per month |
| 116 | - Formula: `Estimated Traffic x Conversion Rate Improvement x Average Customer Value = Monthly ROI` |
| 117 | |
| 118 | **GEO/SEO ROI:** |
| 119 | - AI search visibility can drive 10-30% additional organic traffic |
| 120 | - Featured in AI answers = estimated 5-15% traffic increase |
| 121 | - Schema markup improves click-through rates by 20-30% |
| 122 | - Formula: `Current Organic Traffic x Visibility Improvement % x Customer Value = Monthly ROI` |
| 123 | |
| 124 | **Legal/Compliance ROI:** |
| 125 | - ADA lawsuits average $5,000-$75,000 in settlements |
| 126 | - GDPR fines can reach 4% of annual revenue |
| 127 | - Privacy compliance prevents an average of $50,000 in potential penalties |
| 128 | - Formula: `Risk Avoidance = Probability of Issue x Average Cost` |
| 129 | |
| 130 | **Sales ROI:** |
| 131 | - Improved sales materials increase |