Solution Shaping

Shape source material into an Interaction Tree and State Model — a story map and domain model. Contains 11 rules across 4 categories to guide shaping of requirements into epics, sub-epics, and stories with associated state concepts.

When to Apply

Reference these guidelines when:

• Shaping requirements from source documents into a story map
• Deriving epics and stories from user journeys or business flows
• Building an Interaction Tree (hierarchical actor exchanges)
• Modeling domain state concepts (State Model)
• Defining story granularity and slice order

Rule Categories by Priority

Quick Reference

1. Source & Scope (HIGH)

• source-derive-from-source — Derive concepts from interactions in the source; do not invent workflows
• source-logical-domain-level — Keep at logical/domain level; no implementation details
• source-speculation-assumptions — State assumptions when unclear; do not speculate or invent

2. Hierarchy (HIGH)

• hierarchy-parent-granularity — Keep parent nodes at appropriate granularity; do not leak child detail
• hierarchy-sequential-order — Order tree sequentially; required state creators before consumers

3. State & Structure (HIGH)

• state-required-state — Shared state on parent only; comprehensive preconditions
• state-resulting-state — Same inheritance as required state; outcome language only
• state-failure-modes — Max 3 per interaction; domain rules only
• state-concept-scoping — Scope concepts to Epic/Story that owns them
• state-structured-concepts — Complete full workflow (interactions → concepts → State Model) for each slice

4. Interaction (MEDIUM-HIGH)

• interaction-supporting-actor-response — Supporting = system; Actor → System exchange
• interaction-story-granularity — Break down by distinct areas; sufficient stories for rule detail

How to Use

Read individual rule files for detailed DO/DO NOT guidance:

rules/source-derive-from-source.md
rules/state-required-state.md

Each rule file contains:

• Brief explanation
• DO with example
• DO NOT with example

Full Compiled Document

For the complete guide with all rules, core definitions, output structure, validation, and shaping process: AGENTS.md

Build

To regenerate AGENTS.md from rules and content:

python scripts/build_agents_md.py

Run from the solution-shaping skill directory.