| name | seed-extraction |
| model | opus |
| description | Produces a seed file — a YAML-fronted markdown document capturing a reusable pattern with trigger, evidence, and application steps — from a conversation or project experience. Use when: 'extract the learnings', 'turn this into a seed', 'what can we learn from this', 'capture this insight', 'document this pattern'. |
| category | wisdom-garden |
| triggers | ["extract the learnings","turn this into a seed","what can we learn from this","capture this insight","document this pattern"] |
| tier | 1 |
| agents | ["primary"] |
| tool_dependencies | ["file_system"] |
| inputs | [{"name":"experience","type":"string","description":"Description of the conversation, project, or experience to extract a reusable pattern from","required":true}] |
| outputs | [{"name":"seed_file","type":"ref","format":"cas-ref","description":"Seed file capturing a reusable pattern with trigger, evidence, and application steps"}] |
Seed Reflector Skill
Version: 1.0
Created: 2026-02-02
Author: Manus
Purpose: Extract, document, and apply reusable "seeds" (patterns, insights, principles) from experiences
Overview
This skill encodes the practice of seed extraction and reflection — identifying reusable patterns from experiences and documenting them in a way that makes them easy to apply in future contexts.
Philosophy: Every experience contains seeds. The practice is learning to see them, extract them, and plant them where they'll grow.
When to Use This Skill
- After completing a major project or release
- When you notice a pattern emerging across multiple experiences
- During memory maintenance (Tier A → Tier B compression)
- When preparing to share knowledge with other agents
- When you want to reflect on what you've learned
What Is a "Seed"?
A seed is a reusable pattern, insight, or principle that:
- Emerged from experience (not abstract theory)
- Can be applied in future contexts (not one-time specific)
- Has a clear trigger (you know when to use it)
- Captures wisdom (not just information)
Examples of Seeds:
- "Three-Tiered Governance" (from Dataiku research)
- "Harness Trace" (traceability pattern)
- "Context Iceberg" (hierarchical context management)
- "3-Month Rule" (semantic compression heuristic)
Seed Extraction Process
Step 1: Identify Candidate Patterns
Look for:
- Decisions that worked well (or didn't)
- Patterns that emerged across multiple instances
- Insights that changed how you think
- Principles that guided successful outcomes
- Tensions or tradeoffs you navigated
Questions to ask:
- What did I learn that I didn't know before?
- What pattern did I notice repeating?
- What decision framework did I use?
- What would I do differently next time?
- What would I tell someone else facing this situation?
Step 2: Test for Reusability
A good seed is:
