@arthai/agents 1.0.0

package/skills/pitch-generator/SKILL.md

@@ -0,0 +1,378 @@
+---
+name: pitch-generator
+description: "Generate consulting proposals and pitch documents. Usage: /pitch-generator <client-name> [--format proposal|exec-summary|deck-outline]"
+user-invocable: true
+arguments: "<client-name> [--format proposal|exec-summary|deck-outline]"
+---
+
+# Consulting Pitch & Proposal Generator
+
+Generate polished consulting proposals, executive summaries, or pitch deck outlines from client discovery and opportunity data. Three output formats available, each tailored to a different audience and decision stage.
+
+## Prerequisites
+
+This skill requires completed discovery and opportunity mapping. Check for files at:
+```
+clients/<client-name>/profile.md
+clients/<client-name>/discovery/current-state.md
+clients/<client-name>/discovery/maturity-assessment.md
+clients/<client-name>/assessment/opportunity-matrix.md
+clients/<client-name>/assessment/roadmap.md
+```
+
+If discovery files are missing, redirect:
+```
+Discovery data not found for <client-name>. Run /client-discovery <client-name> first.
+```
+
+If opportunity mapping files are missing, redirect:
+```
+Opportunity data not found for <client-name>. Run /opportunity-map <client-name> first.
+```
+
+## Step 1 — Load Client Data
+
+1. Read all files in `clients/<client-name>/`
+2. Extract key data points:
+   - Company name, industry, size, revenue range
+   - Pain points and goals
+   - AI maturity score and readiness level
+   - Top-ranked initiatives from opportunity matrix
+   - Three-wave roadmap
+   - ROI data (if available from /roi-calculator)
+   - Department recommendations
+
+## Step 2 — Determine Format
+
+Parse the `--format` flag. Default to `proposal` if not specified.
+
+| Format | Audience | Length | Purpose |
+|---|---|---|---|
+| `proposal` | Decision-makers + technical leads | 5-8 pages | Full engagement proposal for budget approval |
+| `exec-summary` | C-suite / board | 1 page | Quick decision document for executive review |
+| `deck-outline` | Presentation audience | 9 slides | Slide-by-slide outline for live pitch |
+
+## Step 3 — Generate Document
+
+### Format: `proposal` (Full Proposal — 5-8 pages)
+
+Generate a complete consulting proposal with the following sections:
+
+#### Section 1: Cover Page
+
+```
+═══════════════════════════════════════════════════════
+          AI TRANSFORMATION PROPOSAL
+
+          Prepared for: [Company Name]
+          Prepared by: [Your Firm / Your Name]
+          Date: [Current Date]
+
+          CONFIDENTIAL
+═══════════════════════════════════════════════════════
+```
+
+#### Section 2: Executive Summary (<400 words)
+
+Write a compelling executive summary that covers:
+- The client's current situation and key challenges (2-3 sentences)
+- The opportunity AI presents for their specific business (2-3 sentences)
+- The recommended approach at a high level (2-3 sentences)
+- Expected outcomes and ROI headline number (2-3 sentences)
+- Call to action (1 sentence)
+
+This section must be under 400 words. It should stand alone — a reader who only reads this section should understand the full value proposition.
+
+#### Section 3: Current State Assessment
+
+Summarize from discovery data:
+- Business context and market position
+- Technology landscape (include Mermaid architecture diagram from current-state.md)
+- AI maturity scorecard (reproduce the ASCII visualization)
+- Key gaps and risks of inaction
+
+#### Section 4: Opportunities Identified
+
+Present the top opportunities:
+- Include the 2x2 priority matrix (ASCII) from opportunity-matrix.md
+- Highlight the top 3-5 initiatives by weighted score
+- For each highlighted initiative: name, description, expected impact, feasibility assessment
+- Connect each opportunity to a specific pain point the client raised
+
+#### Section 5: Recommended Approach
+
+Detail the phased engagement:
+- Overview of the three-wave approach
+- Why phased delivery reduces risk and builds confidence
+- Governance model (steering committee, sprint reviews, success metrics)
+- Our methodology (discovery, pilot, scale framework)
+
+#### Section 6: Solution Architecture
+
+Generate a Mermaid diagram showing the proposed AI-enabled architecture:
+
+```mermaid
+graph TB
+    subgraph "Current State"
+        A[Existing Systems]
+    end
+    subgraph "AI Layer"
+        B[Data Pipeline]
+        C[ML Platform]
+        D[AI Services]
+    end
+    subgraph "Business Impact"
+        E[Automated Workflows]
+        F[Intelligent Insights]
+        G[Predictive Capabilities]
+    end
+    A --> B
+    B --> C
+    C --> D
+    D --> E
+    D --> F
+    D --> G
+```
+
+Customize based on actual recommended initiatives and existing tech stack.
+
+#### Section 7: Implementation Roadmap
+
+Generate a Mermaid Gantt chart showing the three-wave roadmap:
+
+```mermaid
+gantt
+    title AI Transformation Roadmap
+    dateFormat  YYYY-MM-DD
+    axisFormat  %b %Y
+
+    section Wave 1: Quick Wins
+    Initiative 1           :w1a, 2026-04-01, 30d
+    Initiative 2           :w1b, 2026-04-01, 45d
+    Wave 1 Review          :milestone, m1, after w1b, 0d
+
+    section Wave 2: Foundation
+    Initiative 3           :w2a, after m1, 45d
+    Initiative 4           :w2b, after m1, 60d
+    Wave 2 Review          :milestone, m2, after w2b, 0d
+
+    section Wave 3: Strategic
+    Initiative 5           :w3a, after m2, 45d
+    Initiative 6           :w3b, after m2, 60d
+    Final Review           :milestone, m3, after w3b, 0d
+```
+
+Adjust dates, initiative names, and durations based on the actual roadmap.
+
+#### Section 8: Investment & ROI
+
+Present a clear investment table:
+
+```
+Investment Summary
+══════════════════════════════════════════════════════════════════
+                        Wave 1      Wave 2      Wave 3      Total
+──────────────────────────────────────────────────────────────────
+Implementation          $XX,XXX     $XX,XXX     $XX,XXX     $XXX,XXX
+Infrastructure          $X,XXX      $XX,XXX     $XX,XXX     $XX,XXX
+Training & Change Mgmt  $X,XXX      $X,XXX      $X,XXX      $XX,XXX
+──────────────────────────────────────────────────────────────────
+Total Investment        $XX,XXX     $XX,XXX     $XX,XXX     $XXX,XXX
+══════════════════════════════════════════════════════════════════
+
+Expected Annual Return: $XXX,XXX - $X,XXX,XXX
+Payback Period: X-X months
+3-Year ROI: XXX% - XXX%
+```
+
+Include the "cost of doing nothing" — estimate what the client loses each month/year by not adopting AI:
+- Continued manual costs
+- Competitive disadvantage
+- Missed revenue opportunities
+- Talent attrition risk
+
+If ROI calculator data is available (from /roi-calculator), use those numbers. Otherwise, provide conservative estimate ranges based on industry benchmarks.
+
+#### Section 9: Team & Engagement Model
+
+Describe the proposed team structure:
+- Engagement lead (role and responsibilities)
+- Technical architect (role and responsibilities)
+- ML/AI engineers (role and responsibilities)
+- Client-side team requirements (who needs to be involved)
+- Communication cadence (weekly standups, bi-weekly demos, monthly steering)
+
+#### Section 10: Next Steps
+
+Clear call to action:
+1. Schedule a 60-minute deep-dive session to review this proposal
+2. Identify internal champions and steering committee members
+3. Sign engagement letter and SOW for Wave 1
+4. Kick off discovery sprint (Week 1)
+
+Include contact information placeholder and proposed start date.
+
+---
+
+### Format: `exec-summary` (1-Page Executive Summary)
+
+Generate a single-page executive summary with these sections, each kept extremely concise:
+
+```
+═══════════════════════════════════════════════════════
+  AI OPPORTUNITY BRIEF — [Company Name]
+  [Date] | Confidential
+═══════════════════════════════════════════════════════
+
+THE PROBLEM
+[2-3 sentences: What pain points are costing the client money/time/market share]
+
+THE OPPORTUNITY
+[2-3 sentences: What AI can do for this specific business, with one quantified claim]
+
+OUR APPROACH
+[3-4 bullet points: Phased delivery, key initiatives, timeline]
+
+EXPECTED ROI
+[Key metrics: annual savings, revenue impact, payback period — in a mini-table]
+
+INVESTMENT
+[Total range and per-wave breakdown in one line each]
+
+TIMELINE
+[One-line summary: "6-month engagement across 3 waves, first results in 60 days"]
+
+NEXT STEP
+[One sentence CTA: "Schedule a 60-minute deep-dive to review the full proposal."]
+
+═══════════════════════════════════════════════════════
+```
+
+Total length must fit on one printed page (~500 words max).
+
+---
+
+### Format: `deck-outline` (9-Slide Pitch Deck Outline)
+
+Generate a slide-by-slide outline with speaker notes. The deck must map Arth AI's **actual agent ecosystem** to the client's needs — not generic consulting language. Reference specific agents and skills by name.
+
+#### Slide 1: Title
+- "Arth AI x [Client Name]"
+- Subtitle that captures the engagement thesis (not generic "AI Transformation")
+- Date and confidentiality notice
+- Speaker notes: Opening hook — reference a specific pain point from discovery
+
+#### Slide 2: Strategic Context
+- Reframe the client's stated vision into 2-3 strategic dimensions (synthesize, don't parrot)
+- Show the interdependencies between their goals
+- Highlight the key insight: what must happen first before the rest works
+- Speaker notes: Demonstrate that we've analyzed their strategy, not just read it
+
+#### Slide 3: The Arth AI Agent Ecosystem
+- High-level overview of what Arth AI brings: 20 agents + 30 skills, organized by capability category
+- Present as a capability map with 6 categories:
+
+| Category | Agents | What They Do |
+|---|---|---|
+| Software Engineering | `architect`, `code-reviewer`, `frontend`, `python-backend` | Design, build, review, ship |
+| Quality Assurance | `qa` + 5 specialized QA agents | Test strategy, generation, quality gates |
+| Product & Design | `product-manager`, `design-studio` (3 agents) | Ideate, design, critique, plan |
+| Infrastructure & Ops | `sre`, `ops` | Deploy, monitor, incident response |
+| GTM & Sales | `gtm-expert` | Position, enable, compete |
+| Strategy & Consulting | `ai-consultant` | Discover, assess, propose, track |
+
+- Speaker notes: "This isn't a slide deck about AI. These are production agents your team works with daily — they design your systems, review your code, build your products, and run your infrastructure."
+
+#### Slide 4: Department × Agent Mapping
+- **This is the differentiator slide.** Map specific agents and skills to each of the client's departments.
+- Table format:
+
+| Department | Their Ask | Arth AI Agents & Skills | Example Use |
+|---|---|---|---|
+| [Dept 1] | [From discovery] | `agent1`, `agent2`, `/skill1` | [One concrete scenario] |
+| [Dept 2] | [From discovery] | `agent3`, `/skill2`, `/skill3` | [One concrete scenario] |
+
+- Include one "in practice" example per department showing an agent workflow
+- Speaker notes: Walk through each row — "For [department], your team uses the [agent] to [specific action]. Here's what that looks like: [scenario]."
+
+#### Slide 5: Phase 1 — Deploy & Discover (Weeks 1-4)
+- Focus: Deploy quick-win agents + run consulting discovery
+- Table: What We Deploy | What Your Team Gets | Departments
+- Must reference specific agents being deployed (e.g., "Deploy `code-reviewer` and `qa` agents to Engineering")
+- Include the consulting discovery deliverables (`/client-discovery` outputs)
+- Speaker notes: "By week 4, your engineering team is shipping with AI code review, your QA has automated test generation, and we have a full current-state assessment to drive Phase 2."
+
+#### Slide 6: Phase 2 — Build & Innovate (Weeks 5-10)
+- Focus: Use agents to BUILD new AI features and products
+- This is where `product-manager`, `design-studio`, `architect`, `frontend`, `python-backend` agents come in
+- Show the product development workflow: `product-manager` → `design-studio-think` → `design-studio-create` → `architect` → `frontend`/`python-backend` → `code-reviewer` → `qa`
+- The client's team works alongside these agents to ideate, design, and build their AI product features
+- Include `/planning` and `/implement` skills
+- Speaker notes: "Phase 2 is where your product vision becomes real. Your product leads use the `product-manager` agent to scope features, `design-studio` to iterate on UX, and our engineering agents to build and ship. Your team learns the methodology by working inside it."
+
+#### Slide 7: Phase 3 — GTM & Scale (Weeks 11-16)
+- Focus: Package capabilities into sellable services + methodology transfer
+- Deploy `gtm-expert` for market positioning and sales enablement
+- Use `/deliverable-builder` for case studies, board decks, training plans
+- Use `/market-research` for competitive positioning
+- Methodology transfer: the client's team inherits the agent ecosystem and consulting playbooks
+- Speaker notes: "Phase 3 converts everything you've built into revenue. The `gtm-expert` agent positions your AI services, builds your sales collateral, and analyzes your competitive landscape."
+
+#### Slide 8: What We Need to Explore Together
+- Discovery inputs required to move from proposal to SOW
+- Table: Category | What We Need | Why It Matters
+- Speaker notes: "These are the inputs that transform this directional strategy into a costed execution plan."
+
+#### Slide 9: Next Steps
+- 3 clear action items with owners
+- Commitment statement: "Build capability, not dependency"
+- Contact information
+- Speaker notes: End with the autonomy pitch — "Our success is measured by how quickly you don't need us."
+
+## Step 4 — Quality Gate
+
+Before outputting the final document, run these quality checks:
+
+1. **Fact Check**: Every claim, metric, and data point must trace back to client-provided materials, a cited industry benchmark, or a verified source. If a number cannot be sourced, remove it or mark it `[TO BE VALIDATED]`. Never fabricate statistics, dollar figures, percentages, or timelines.
+2. **Consistency Check**: All numbers, dates, and initiative names match across sections
+3. **Pain Point Coverage**: Every client pain point from discovery is addressed at least once
+4. **Specificity Check**: No generic placeholder language remains — all content is customized to the client
+5. **Length Check**: Proposal is 5-8 pages, exec-summary is under 500 words, deck-outline has exactly 9 slides
+6. **CTA Check**: Clear next steps with specific actions are included
+7. **Tone Check**: Professional consultant language — measured, authoritative, non-committal on unvalidated specifics. Use qualifying language for unconfirmed items ("initial assessment indicates," "subject to validation," "pending discovery"). Never salesy or overpromising.
+8. **Data Check**: All ROI figures are reasonable, defensible, and sourced. If ROI data is not available, state the opportunity qualitatively and note that quantification requires a discovery deep-dive with client data.
+9. **Source Attribution**: Client-provided information must be attributed ("Per your AI-First Strategy presentation," "As outlined in your roadmap"). This builds trust and demonstrates grounding in the client's reality.
+10. **Open Questions**: Unknown items must be framed as collaborative discovery questions, not filled with assumptions. Include a clear "What we need to explore together" section.
+
+If any check fails, fix the issue before outputting.
+
+## Step 5 — Output Files
+
+Save the generated document(s) in `clients/<client-name>/proposals/`:
+
+| Format | Output File |
+|---|---|
+| `proposal` | `proposals/full-proposal.md` |
+| `exec-summary` | `proposals/executive-summary.md` |
+| `deck-outline` | `proposals/deck-outline.md` |
+
+If the `proposals/` directory doesn't exist, create it.
+
+After saving, inform the user:
+```
+Proposal generated: clients/<client-name>/proposals/[filename].md
+
+To generate other formats, run:
+  /pitch-generator <client-name> --format exec-summary
+  /pitch-generator <client-name> --format deck-outline
+```
+
+## Notes
+
+- Never use generic consulting language — every sentence must reference the client's specific situation
+- ROI projections should be conservative — use ranges rather than point estimates
+- If ROI calculator has been run, use those numbers; otherwise, use industry benchmark ranges
+- Mermaid diagrams must be syntactically valid and renderable
+- The proposal should read as if written by a senior consultant, not generated by AI
+- Include "[Your Firm Name]" as a placeholder where the consulting firm name should go
+- Flag any sections where data is insufficient with "[DATA NEEDED: description]" markers