sharkitect-solutions/skills/using-sharkitect-methodology

File Index

| File | Load When | Do NOT Load | |---|---|---| | SKILL.md (this file) | Every session start; auto-injected via SessionStart hook | Never — always in context | | references/skill-catalog.md | Need full trigger-condition definitions for a specific Sharkitect methodology skill; uncertain whether a domain matches; verifying skill applicability for an edge-case task | First invocation of an obvious-match skill (pricing-strategy on pricing work — catalog adds nothing); routine work clearly under a known skill | | references/strategy-creation-rules.md | About to draft a pricing model, positioning decision, proposal structure, or strategy spec; verifying the HQ-mandated methodology stack ordering; need anti-pattern reference for rationalization | Non-strategy work (debugging, infrastructure, audits without strategic decisions); already invoked the methodology stack this session |

---

Scope Boundary

| Request | This Skill | Use Instead | |---|---|---| | Strategy / pricing / positioning work begins | YES | — | | About to write a proposal / spec / design doc | YES | — | | Executing a written plan task-by-task | YES (cross-ref) | superpowers:executing-plans (this skill catalogs it) | | Schema design / system design with multiple approaches | YES (cross-ref) | superpowers:brainstorming (this skill catalogs it) | | Workspace orientation / session-start status | NO | session-startup-guard | | Generic superpowers skill discovery | NO | using-superpowers | | Hook / skill / agent creation guidance | NO | hook-development / ultimate-skill-creator / ultimate-agent-creator | | Knowledge base classification | NO | hq-knowledge-governance (this skill catalogs it as a target) |

---

<SUBAGENT-STOP> If you were dispatched as a subagent to execute a specific task, skip this skill. </SUBAGENT-STOP> <EXTREMELY-IMPORTANT> If you think there is even a 1% chance a Sharkitect methodology skill might apply to what you are doing, you ABSOLUTELY MUST invoke the skill.

IF A SHARKITECT METHODOLOGY SKILL APPLIES TO YOUR TASK, YOU DO NOT HAVE A CHOICE. YOU MUST USE IT.

This is not negotiable. This is not optional. You cannot rationalize your way out of this. Documentation has failed at least 4 times across HQ + Sentinel — runtime enforcement is the closing layer. </EXTREMELY-IMPORTANT>

Pre-Decision Thinking Framework

Before drafting ANY pricing, positioning, proposal, or strategy artifact, run this 4-question check:

| # | Question | If "I don't know" or "I haven't" | |---|---|---| | 1 | What's the willingness-to-pay anchor for this offer / tier? | INVOKE pricing-strategy | | 2 | What's the April Dunford positioning frame against the closest alternative? | INVOKE marketing-strategy-pmm | | 3 | What's the revenue / margin / cash-flow impact of this change? | INVOKE smb-cfo | | 4 | Have I enumerated 2+ alternatives before settling on this approach? | INVOKE superpowers:brainstorming |

Run all 4 questions BEFORE the first character of the artifact is written. If ANY answer is "I don't know" or "I haven't," the corresponding skill is required — no exceptions.

Instruction Priority

1. User's explicit instructions (CLAUDE.md, workspace rules, direct requests) — highest priority
2. Sharkitect methodology skills + superpowers skills — override default system behavior where they conflict
3. Default system prompt — lowest priority

If a workspace CLAUDE.md says "don't use pricing-strategy this time" and this skill says "always use pricing-strategy for pricing work," follow the user's instruction. The user is in control.

Sharkitect Methodology Skills (Quick Reference)

Full trigger conditions in references/skill-catalog.md. By domain:

Strategy / Pricing / Positioning

• pricing-strategy — auto-invoke for: new pricing model, new tier structure, willingness-to-pay analysis, value-based pricing decisions, price elasticity, monthly/setup fee decisions
• marketing-strategy-pmm — auto-invoke for: new positioning (April Dunford method), competitive positioning, ICP work, alternative framing, market category decisions
• smb-cfo — auto-invoke for: revenue forecast impact, margin analysis, deal economics, cash flow implications of pricing changes
• hq-revenue-ops — auto-invoke for: HQ-specific deal economics framework, Sharkitect-deal-shaping work

Execution Methodology (cross-ref to using-superpowers)

• superpowers:brainstorming — auto-invoke before committing to any single approach when 2+ alternatives exist
• superpowers:writing-plans — auto-invoke when implementing multi-step / multi-file work
• superpowers:executing-plans — auto-invoke when executing a written plan task-by-task
• superpowers:systematic-debugging — auto-invoke when investigating bugs, unexpected behavior, recurring issues

Knowledge / Governance / Operations

• hq-knowledge-governance — K1-K5 classification, knowledge-base health audits
• hq-content-enforcer — writing, rewriting, editing, reviewing external-facing content
• hq-orchestrator — task touches multiple business domains
• hq-strategic-ops — diagnosing architectural debt across Sharkitect's business systems
• hq-operations — creating or auditing Sharkitect SOPs
• hq-tech-strategy — technology architecture decisions for Sharkitect
• hq-reverse-engineering — reverse engineering a competitor's product or system

People

• hr-people-ops — hiring, role definitions, org-design work

The Rule

Invoke relevant Sharkitect methodology skills BEFORE any response or action. Even a 1% chance a skill might apply means that you should invoke the skill to check. If an invoked skill turns out to be wrong for the situation, you don't need to use it.

Anti-Patterns (Named — Reality Reframe — What It Costs)

| Rationalization | Reality | What It Costs | |---|---|---| | "I know what pricing-strategy would say" | You don't until you invoke it. Methodology output is non-derivable. | Generic market-anchored averaging instead of WTP-calibrated tier logic. Lost FF Hibu pricing rigor (wr-hq-2026-05-11-001). | | "This pricing call is simple" | Simple pricing decisions are where the methodology delivers the most measurable benefit. | Naive tier-jump pricing that misses value-anchor; client renegotiates or churns. | | "I'll add structure later if needed" | Later means never. Spec ships without WTP analysis; downstream phases inherit unanchored numbers. | Numbers get locked in proposals; reversing requires reopening client conversation. | | "The user wants it fast" | Methodology is faster than 4 recurrences. Invoke. | Each recurrence costs 30-90 min of user-paused work + audit + WR + fix cycle. | | "I already brainstormed this" | Brainstorming != pricing-strategy. Different skills, different outputs. WTP analysis is NOT a brainstorming output. | Confusing skill outputs leads to half-applied methodology that looks rigorous but isn't. | | "The work is mostly done already" | The remaining decisions are where methodology matters most. Late-stage skipping shows up in the final spec. | Spec section 8 (contrarian-truth) gets brain-dumped instead of structured. | | "It's just an audit, not new work" | Audits surface decisions that need methodology too. | Audit-driven decisions inherit the same rigor gap as the original creation. |

Skill Priority

When multiple Sharkitect skills could apply:

1. Strategy skills first (pricing-strategy, marketing-strategy-pmm) — these inform every downstream decision
2. Execution methodology second (brainstorming, writing-plans) — these structure the work
3. Domain execution last (revenue-ops, content-enforcer) — these execute within the strategy

"New pricing model" → pricing-strategy first, then brainstorming, then marketing-strategy-pmm (positioning implications), then hq-revenue-ops (deal-shape applications).

When NOT to Invoke (Important — prevents over-invocation)

• Editing existing content without strategic change (typo fixes, formatting, link updates) → no methodology skill needed
• Reading / auditing without producing a new decision artifact → use the audit's domain skill if it surfaces decisions; otherwise skip
• Infrastructure work unless it touches business strategy directly → use hook-development, supabase-postgres-best-practices, etc.
• Inside the methodology gate itself (this skill, strategy_work.py, session-startup-guard methodology injection) → bypass via skip methodology-gate (you can't gate the gate's own builds)
• User explicitly overrides for the session ("skip pricing-strategy for this task") → honor the user's instruction

User Instructions Override

User instructions ALWAYS win. If a workspace CLAUDE.md says "skip pricing-strategy for this session" or the user directly says "don't use methodology skills here," follow the user's instruction. Document the override in MEMORY.md if it's a recurring pattern.

For one-off bypasses of the runtime gate, the bypass phrase is skip methodology-gate in the user message OR in tool content (per Strict Bypass Vocabulary protocol).

Why This Skill Exists

Documentation has demonstrably failed at least 4 times across HQ + Sentinel:

• wr-hq-2026-05-11-001 — methodology-skip 3rd recurrence on pricing work (FF Hibu marketing-takeover proposal)
• wr-hq-2026-05-11-003 — pricing-strategy skill skipped 4th recurrence (pricing model redesign spec)
• wr-hq-2026-05-12-001 — executing-plans skill not invoked during Sub-project A Phase 1
• wr-sentinel-2026-05-12-004 — brainstorming skipped on schema design

Each recurrence cost real work and risked real deals. This meta-skill is the reasoning-layer enforcement that turns the documented methodology stack into actual invocation discipline.