examples / example-04-feature-design-with-arena.md
examples / example-04-feature-design-with-arena.md
A team needs to design a replaceable index-reading interface for article search. The interface must support today's static JSON config and a future remote source. The team wants to compare multiple approaches before committing to one.
This maps to the Feature playbook combined with the Arena skill. The real pstack pattern is: describe the requirement, let pstack route it to Feature, then explicitly ask for an arena comparison of structurally different approaches.
Step 1: Route to the Feature playbook
{ "tool": "playbook-route", "parameters": { "task": "Design a replaceable index-reading interface for article search. It must support today's static JSON config and a future remote source. Compare at least two structurally different approaches.", "limit": 1 } }
Step 2: Load the Feature playbook
Step 3: Get the Arena skill prompt for comparison
Step 4: Get the principle about exhaustive design space
Step 5: Use the skill-graph to check dependencies
pstack-mcp returns:
The model produces a design comparison table with trade-offs, then selects the best approach and writes the interface spec before any code.
A team needs to design a replaceable index-reading interface for article search. The interface must support today's static JSON config and a future remote source. The team wants to compare multiple approaches before committing to one.
This maps to the Feature playbook combined with the Arena skill. The real pstack pattern is: describe the requirement, let pstack route it to Feature, then explicitly ask for an arena comparison of structurally different approaches.
Step 1: Route to the Feature playbook
{ "tool": "playbook-route", "parameters": { "task": "Design a replaceable index-reading interface for article search. It must support today's static JSON config and a future remote source. Compare at least two structurally different approaches.", "limit": 1 } }
Step 2: Load the Feature playbook
Step 3: Get the Arena skill prompt for comparison
Step 4: Get the principle about exhaustive design space
Step 5: Use the skill-graph to check dependencies
pstack-mcp returns:
The model produces a design comparison table with trade-offs, then selects the best approach and writes the interface spec before any code.
{
"tool": "get_playbook",
"parameters": {
"name": "feature",
"mode": "full"
}
}
{
"tool": "skill-prompt",
"parameters": {
"skill_name": "arena",
"task_context": "pit three approaches: (1) strategy pattern with registry, (2) plugin interface with dynamic loading, (3) configuration-driven adapter. Evaluate each on extensibility, complexity, and performance."
}
}
{
"tool": "get_principles",
"parameters": {
"names": ["exhaust-the-design-space", "model-the-domain"],
"mode": "full"
}
}
{
"tool": "skill-graph",
"parameters": {
"skill_name": "architect"
}
}
{
"tool": "get_playbook",
"parameters": {
"name": "feature",
"mode": "full"
}
}
{
"tool": "skill-prompt",
"parameters": {
"skill_name": "arena",
"task_context": "pit three approaches: (1) strategy pattern with registry, (2) plugin interface with dynamic loading, (3) configuration-driven adapter. Evaluate each on extensibility, complexity, and performance."
}
}
{
"tool": "get_principles",
"parameters": {
"names": ["exhaust-the-design-space", "model-the-domain"],
"mode": "full"
}
}
{
"tool": "skill-graph",
"parameters": {
"skill_name": "architect"
}
}