Offering Innovation

A Claude Code plugin that transforms a problem statement into a market offering design — grounded in jobs theory, first principles decomposition, and competitive positioning.

Run Phase 1 on this problem: [your problem statement]

The Problem It Solves

Most offering development starts too late. Teams jump from "here's a problem" to "here's a solution" without understanding who actually has the problem, what they're fundamentally trying to accomplish, or why existing solutions fall short. The result is an offering that solves a symptom, targets the wrong buyer, or lands in a competitive space with no differentiation.

This plugin forces the work that gets skipped: atomic root cause analysis, stakeholder system mapping, ODI-validated job mapping, and a diverge-then-converge offering design with explicit competitive positioning. Every claim requires evidence. Every capability traces to an unmet outcome.

How It Works

Five phases run in sequence. Each phase ends with a gate before you proceed.

| Phase | What It Does | Key Output | |-------|-------------|-----------| | 1. Problem Decomposition | Strip the problem to irreducible root causes via first principles and 5 Whys | Atomic problem statement + execution barriers + unmet metrics | | 2. System Mapping | Identify all stakeholders, classify job executors, map cross-dependencies | Stakeholder map + ranked executors | | 3. Job Mapping | Build an ODI-compliant job map per executor | Core job + 8-step map + validated outcome statements | | 4. Offering Design | Diverge (unconstrained ideation) then converge (competitive analysis) | Offering definition + capabilities + proof points | | 5. Analysis & Executive Summary | Synthesize the full pipeline into a decision-ready document | Chain-of-logic executive summary + validation plan |

Installation

Requirements: Claude Code, Node.js

npx offering-innovation-plugin

Run outputs save to runs/ in the current folder. Open that folder in Claude Code and tell Claude your problem.

Global install (commands and skills available in all projects):

npx offering-innovation-plugin --global

OpenCode

cp -r dist/opencode/. your-project/

GitHub Copilot

cp -r dist/github/.github your-project/

Gemini CLI

cp -r dist/gemini/.gemini your-project/

Codex CLI

cp -r dist/codex/.agents your-project/
mkdir -p your-project/.codex
cp -r dist/codex/.codex/agents your-project/.codex/

Trae

cp -r dist/trae/.trae/skills/* ~/.trae/skills/

Rovo Dev

# Project-specific
cp -r dist/rovo-dev/.rovodev your-project/

# Or global
cp -r dist/rovo-dev/.rovodev/skills/* ~/.rovodev/skills/

Qoder

# Project-specific
cp -r dist/qoder/.qoder your-project/

# Or global
cp -r dist/qoder/.qoder/skills/* ~/.qoder/skills/

Pi

cp -r dist/pi/.pi your-project/

Usage

Copy runs/_template/ to start a new run:

cp -r runs/_template/ runs/run_001_your-topic/

Then tell Claude your problem:

Run Phase 1 on this problem: [your problem statement]

Or with an existing run folder open, use /init to be guided through setup.

What One Run Produces

| Artifact | What It Contains | |----------|-----------------| | phase-1-output.md | Root-cause-grounded problem statement, evidence-backed, solution-agnostic | | phase-2-output.md | All actors classified, cross-dependencies mapped, executors ranked | | phase-3-output.md | Core job, 8-step job map, validated outcome statements with importance signals | | phase-4-output.md | Offering definition, capabilities, competitive positioning, falsifiable proof points | | executive-summary.md | Chain of logic, deep offering description, validation plan, net assessment |

Methodology

| Approach | Where Used | Why | |----------|-----------|-----| | First Principles Decomposition | Phase 1 | Separates structural truths from inherited assumptions | | 5 Whys | Phase 1 | Reaches terminal root causes; citation required at each Why | | ODI (Outcome-Driven Innovation) | Phases 3–4 | Maps directly to measurable executor outcomes; prevents solution-shaped job descriptions | | Design Thinking (Diverge/Converge) | Phase 4 | Unconstrained ideation first; competitive analysis informs selection |

The data model: All outputs map to a 3-tier × 4-domain grid (Job/Problem/Offering/Process at the tier level; JTBD/Problem/Offering/Process at the domain level). This structure ensures every capability traces to an unmet outcome and every outcome traces to a root-cause problem.

Five analytical sub-agents activate at key moments: Assumption Challenger, ODI Compliance Checker, Outcome Research Validator, Cross-Reference Mapper, and Systems Dynamics Checker. These run automatically — you don't invoke them manually.

Validation

Run deterministic checks after each phase:

python tests/validate.py runs/your-run/ --phase 1
python tests/validate.py runs/your-run/ --phase 3
python tests/validate.py runs/your-run/

Results: PASS / WARN / FAIL / SKIP per check. Checks include citation counts, ODI outcome format compliance, solution-agnostic language detection, required sections, and placeholder detection.

Guardrails

• Citations are not optional. The 3-source requirement at each gate exists because analytically constructed problems and jobs produce phantom offerings. Evidence grounds the work.
• One offering by default. Splitting into multiple offerings requires affirmative evidence: different buyers, different value props, or independent adoption paths. The burden of proof is on the split.
• Gate failures block, not warn. A failed gate criterion must be corrected before proceeding — it can't be noted and bypassed.
• The "So What?" test. If you can't name a competitor and explain specifically what you do that they don't, the offering isn't differentiated yet.

License

MIT