mold

Convert fuzzy product ideas into approved design specifications via a two-key handshake.

15|1|Updated May 4, 2026

Search Tags:#API design #spec generation #workflow planning #approval gate #interface contracts #evidence grounding

Ready to run

npx skills add https://github.com/paulnsorensen/easy-cheese --skill mold

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: mold
Source: https://github.com/paulnsorensen/easy-cheese/tree/main/skills/mold
Command: npx skills add https://github.com/paulnsorensen/easy-cheese --skill mold

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Mold converts vague product or engineering intent into a grounded, interface-locked spec that is only written after an explicit two-key approval gate.

Core Features & Use Cases

Iterative design dialogue from fuzzy input: Runs a guided Explore → Ground → Shape → Sketch → Grill → Diagnose flow to clarify requirements, resolve contradictions, and stress-test choices for a durable spec artifact.
Evidence-first grounding: Grounds critical claims using cheez-search/cheez-read and Validate Cycles before proposing interface or decision changes.
Interface and blast-radius control: Uses a shape check to lock public seams via pseudocode signatures and bound upstream/downstream impact before extraction.
Two-key handshake extraction: Requires an explicit user verb plus an agent coherence self-check to prevent drift and silent agent-introduced scope.
Curdled spec + recommended next action: Writes the approved spec to the durable corpus via the resolver and computes the best downstream skill (cook, ultracook, or cheese-factory) before prompting handoff.

Quick Start

Ask the AI to run mold on your fuzzy feature idea and be ready to approve the handshake before the spec is extracted to a durable spec file.

Frequently Asked Questions about mold

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I turn fuzzy product ideas into an approved API design specification?▼

Turning fuzzy ideas into an approved API design spec requires an iterative dialogue that clarifies requirements, resolves contradictions, and locks interface seams before writing artifacts. Evidence grounding stress-tests choices to produce a durable, downstream-safe specification.

What is the best way to lock interface contracts before generating a spec file?▼

The best way to lock interface contracts before spec generation is running a shape check using pseudocode signatures to bound upstream and downstream blast-radius impact. This secures public seams before deterministic artifact extraction occurs.

How does a two-key handshake approval gate work for interface contract generation?▼

A two-key handshake approval gate for spec generation requires an explicit user verb alongside an agent coherence self-check. This mechanism prevents scope drift and silent agent-introduced changes before extracting a deterministic spec file.

Can I use iterative design dialogue to plan workflow planning and scope API features?▼

Yes, iterative design dialogue handles workflow planning and API feature scoping via a guided Explore, Ground, Shape, Sketch, Grill, and Diagnose flow. This process grounds critical claims and resolves contradictions for durable downstream scoping.

Does spec generation require evidence grounding for workflow planning?▼

Spec generation requires evidence grounding for workflow planning to validate critical claims before proposing interface or decision changes. It uses search and validate cycles to ensure API design remains evidence-first and downstream-safe.

What are the limitations of generating specs from fuzzy ideas without interface lockdown?▼

Generating specs from fuzzy ideas without interface lockdown risks unbounded upstream and downstream blast-radius impact, scope drift, and silent agent-introduced scope. Without explicit approval gates, extracted interface contracts lack downstream-safe scoping.