/synthesize-knowledge
Knowledge Synthesizer
Extract and synthesize 10-100 documents at 10x speed
Overview
Runs 4 parallel extraction agents (concepts, evidence, relationships, metadata) across a document corpus, then assembles a synthesis document with full source traceability and 3-level quality gates. Handles file placement, README updates, and cross-linking automatically.
What It Does
- Discovers and classifies source documents, estimates extraction time
- Runs 4 parallel agents: Concept Extractor, Evidence Extractor, Relationship Mapper, Metadata Harvester
- Assembles synthesis with domain-specific templates and auto-generated frontmatter
- Validates through 3 quality gates (extraction, synthesis, integration)
- Places files in correct folders, updates READMEs, creates cross-links
Inputs
- Source folder path
- Target domain
- Optional workstream and quality gate level
Outputs
- Synthesis markdown with frontmatter
- Updated parent README
- Cross-links
- Git commit
Example
You have 12 customer call transcripts scattered across folders. Running /synthesize-knowledge produces a synthesis document with extracted pain points, MEDDPICC fields, and evidence-backed patterns in 23 minutes instead of 6 hours — the extraction engine at the heart of the three-layer architecture.
Deep Dives
Related Skills & Workflows
Prompt Creator
Systematic design for production-quality AI prompts
SkillProduce Content
One idea into blog, newsletter, social, images
SkillDeep Planning
Complex task into step-by-step execution plan
WorkflowNewsletter Production
Multi-agent pipeline producing 3 editions per week
WorkflowCampaign Planning
From brief to fully scoped campaign plan in 20 minutes
WorkflowContent Calendar Builder
Quarterly content calendar grounded in keyword data and ICP pain points
Ready to use /synthesize-knowledge?
This skill ships with every Knowledge OS installation. Set up your system in 90 minutes.
Built and maintained by Victor Sowers at STEEPWORKS