distilling

Analyze a user's canvas state and output cognition sidecar YAML.

Updated Feb 17, 2026

Search Tags:#synthesis #canvas #analyze #steering #hypotheses #fears

Ready to run

npx skills add https://github.com/0xHoneyJar/construct-observer --skill distilling

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: distilling
Source: https://github.com/0xHoneyJar/construct-observer/tree/main/skills/distilling
Command: npx skills add https://github.com/0xHoneyJar/construct-observer --skill distilling

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Distill raw user canvases into structured fears, steering targets, and synthesis features to guide decisions and conversations.

Core Features & Use Cases

Extract hypotheses, gaps, and growth signals from a canvas to surface actionable insights.
Generate minimal cognition sidecars for product and research handoffs.
Apply to research canvases, onboarding feedback, and growth experiments to distill narratives into decisions.

Quick Start

Feed the latest user canvas and growth state to generate a cognition sidecar.

Frequently Asked Questions about distilling

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

FAQPage Schema

How do I extract structured hypotheses and growth signals from raw user canvas markdown?▼

Extract structured hypotheses and growth signals from raw user canvas markdown by analyzing the canvas state, growth state, and provenance records to generate structured cognition output. This distillation process surfaces actionable insights by converting raw narratives into a structured YAML format.

What is a cognition sidecar and when do I need to generate one for product handoffs?▼

A cognition sidecar is a structured YAML document containing fears, steering targets, and synthesis features distilled from user canvases. You need to generate one for product and research handoffs to provide minimal, actionable context that guides decisions and conversations without requiring the full canvas.

How do I analyze canvas state to surface steering targets and synthesis features?▼

Analyze canvas state by consuming inputs like canvas markdown, growth state, provenance records, and configuration to produce structured cognition output. The analysis extracts hypotheses, gaps, and growth signals, returning a YAML document that validates against the cognition sidecar schema with defined fears and steering targets.

Can I use canvas markdown and provenance records to identify user fears and gaps in onboarding feedback?▼

Yes, you can use canvas markdown and provenance records to identify user fears and gaps in onboarding feedback. The skill analyzes these inputs to extract hypotheses and growth signals, distilling the onboarding narratives into structured fears and steering targets for research handoffs.

What is the best way to distill research canvases into structured YAML for growth experiments?▼

The best way to distill research canvases into structured YAML for growth experiments is to feed the latest user canvas and growth state into an analysis process that outputs a single YAML document. This document includes schema version, user data, fears, steering targets, and synthesis features.

Does the cognition sidecar schema require specific fields like synthesis features and steering targets?▼

Yes, the cognition sidecar schema requires specific fields including schema version, user, fears, steering targets, and synthesis features. The YAML output must validate against this schema to ensure the structured cognition contains all necessary elements for guiding decisions.