csuite

# C-Suite Decision Support

Umbrella skill for all executive decision-making: CEO-level strategy, CTO-level technology choices, CMO-level growth planning, competitive intelligence, and project evaluation. Each domain loads its own reference files on demand -- this skill detects the mode, loads the right references, and executes the appropriate framework.

**Scope**: Business decisions with meaningful consequences. Use decision-helper for technical architecture micro-choices, domain agents for code, voice-writer for content, and systematic-debugging for debugging.

---

## Mode Detection

Classify the user's request into exactly one mode before proceeding. If the request spans multiple modes, choose the primary one and note the secondary.

| Mode | Signal Phrases | Role Lens |
|------|---------------|-----------|
| **STRATEGY** | Market entry, partnerships, resource allocation, opportunity, "should I/we", strategic pivots, investment | CEO |
| **TECHNOLOGY** | Build vs buy, vendor, SaaS, tech stack, architecture, adopt, technology choice | CTO |
| **GROWTH** | Content strategy, audience, SEO, marketing, brand, community, positioning, channel | CMO |
| **COMPETITIVE** | Competitor, competition, market landscape, differentiation, positioning against, market share | Cross-role |
| **EVALUATION** | Feasibility, effort estimate, ROI, priority, go/no-go, viability, "is it worth it" | Cross-role |

---

## Reference Loading Table

Load references based on the detected mode. Load only the references required by the mode.

| Signal | Mode | Reference |
|--------|------|-----------|
| Market entry, partnerships, resource allocation, opportunity | STRATEGY | `references/strategic-frameworks.md`, `references/decision-matrices.md` |
| Build vs buy, vendor, SaaS, tech stack, architecture | TECHNOLOGY | `references/tco-framework.md`, `references/vendor-evaluation.md` |
| Content, audience, SEO, marketing, brand, community | GROWTH | `references/audience-segmentation.md`, `references/channel-evaluation.md` |
| Competitor, market landscape, positioning, differentiation | COMPETITIVE | `references/competitive-mapping.md`, `references/market-positioning.md` |
| Feasibility, effort, ROI, priority, go/no-go | EVALUATION | `references/feasibility-scoring.md`, `references/roi-frameworks.md` |

---

## Instructions

### Mode: STRATEGY (CEO)

**Framework**: FRAME -> ANALYZE -> DECIDE

**Phase 1: FRAME** -- Convert the user's question into a structured decision with clear stakes and timeline.

- Name the actual decision (users present symptoms; the real decision is broader)
- Identify irreversibility -- reversible decisions deserve less analysis
- Set a time horizon -- 3-month and 3-year decisions need different frameworks
- Classify the decision type: Expansion, Partnership, Allocation, Pivot, or Timing
- Get the user to state: options (2-4), default path risk, deadline, and what makes it hard

**Gate**: Decision framed as one sentence. Options listed (2-4). Type classified.

**Phase 2: ANALYZE** -- Evaluate each option through multiple lenses with evidence.

For each option, assess: Upside (best realistic + expected outcome), Downside (worst realistic + recovery path + irreversible losses), Requirements (resources, assumptions, dependencies), Opportunity Cost (what you cannot do).

Separate facts from assumptions. Quantify where possible. Load reference files for scoring matrices and strategic frameworks.

**Gate**: All options analyzed. Facts and assumptions labeled. Opportunity costs explicit.

**Phase 3: DECIDE** -- Synthesize into a clear recommendation.

- Apply the reversibility test: one-way doors need high confidence; two-way doors can act faster with a checkpoint
- Produce: Recommendation (one sentence), Confidence (High/Medium/Low), Why this option (2-3 reasons), What must be true (invalidating assumptions), First move (48-hour action), Revisit trigger
- State explicitly what would change the recommendation

**Gate**: Recommendation stated. First action identified. Revisit trigger set.

---

### Mode: TECHNOLOGY (CTO)

**Framework**: SCOPE -> EVALUATE -> RECOMMEND

**Phase 1: SCOPE** -- Define the capability needed, stripped of solution bias.

- Start with the need, not the product ("we need reliable async delivery" not "we need Kafka")
- Quantify hard requirements (latency, throughput, compliance)
- Identify the real driver (build vs buy is sometimes "convince management" or "hire someone")
- List actual options: build from scratch, build on OSS, buy SaaS, buy + customize, do nothing

**Gate**: Capability defined without solution bias. Options enumerated. Hard requirements quantified.

**Phase 2: EVALUATE** -- Score options on dimensions that matter for technology decisions.

- Total cost of ownership at Year 3, not sticker price (the "free" OSS needing a full-time engineer is expensive)
- Score on: Fit (5), TCO (4), Operational burden (4), Team capability (3), Lock-in risk (3), Time to value (3), Flexibility (2)
- Apply the build-vs-buy heuristic: core competency, requirements stability, team capacity, timeline, scale, compliance

Load `references/tco-framework.md` for TCO templates and `references/vendor-evaluation.md` for vendor scorecards.

**Gate**: TCO estimated. Dimensions scored. Build-vs-buy heuristic applied.

**Phase 3: RECOMMEND** -- Deliver a clear recommendation with reasoning.

- Present the weighted scoring matrix
- State: Decision, Confidence, Why this option, Watch-for risks, Migration path, First step
- Define exit criteria: when to reconsider for each option type

**Gate**: Recommendation stated. Exit criteria defined. First step identified.

---

### Mode: GROWTH (CMO)

**Framework**: ASSESS -> STRATEGIZE -> PLAN

**Phase 1: ASSESS** -- Understand current state before recommending.

- Audit: publications, content volume, existing audience, active channels, performance data
- Identify the binding constraint: Discovery, Content, Conversion, Retention, or Capacity
- Creator capacity