| name | skill80-20-knowledge-engine |
| description | Apply the Pareto 80/20 principle to rapidly distill the vital 20% of high-value knowledge, patterns, signals, and insights from large data traces such as X post histories, conversation threads, event logs, recommendation signals, or document corpora. Activate on signal tracing tasks, knowledge extraction for intrinsic pursuit, pattern emergence detection, or when exhaustive review would induce entropy overload. Supports efficient synthesis aligned to core objectives and integrates with evt-processor, labyrinth-os-constitutional-enforcer, unifying-thread, and sovereign PQC/ZK layers. |
Skill 80/20 Knowledge Engine
Rapidly extract the highest-leverage 20% of signal from overwhelming data volumes while preserving structural invariants. Designed for sovereign AI workflows where entropy management and thermodynamic efficiency are non-negotiable.
When to Activate
- Processing large X/Twitter post histories, emoji trails, or recommendation signal streams (follows, likes, impressions, profile views).
- Distilling conversation threads or event logs into actionable patterns.
- Knowledge extraction tasks where full review would create excessive cognitive or computational load.
- Pattern emergence detection in watcher signals, labyrinth corridors, or multi-agent execution traces.
- Integration points with unifying-thread (cross-domain synthesis), evt-processor (structured output), or PQC/ZK graph intelligence layers (private signal distillation).
Core Workflow
- Ingest trace — Accept raw data (posts, signals, logs, documents).
- Apply Pareto filter — Identify the vital 20% using frequency, impact, recurrence, or structural centrality heuristics.
- Extract invariants & patterns — Surface recurring motifs, high-signal nodes/edges, and low-entropy representations.
- Synthesize & map — Align distilled output to user goal hierarchy (hard invariants + soft objectives).
- Output in evt- format — Structured, auditable, machine-readable record for downstream agents and observability.
- Optional ZK/PQC layer — When operating on sensitive graph signals, support private distillation via ZK graph proofs + post-quantum attestation.
Integration with Sovereign Stack
- evt-processor: All outputs are evt- JSON records (schema_version, record_type: "80_20_distillation", tags, connections, payload).
- unifying-thread: Feeds distilled signals into cross-domain unification analysis.
- a2a-skill-registry-manager & github-deployment-auto-scaler: Register and version this skill; monitor its deployment health.
- PQC / ZK graph layer: Combine with ML-DSA/Falcon signatures and ZK proofs for private, quantum-resistant signal processing (e.g., distilling recommendation graph properties without revealing raw edges).
- Genesis Conductor / MCP: Thermodynamic-aware execution (Landauer efficiency via aggressive pruning of low-value signal).
- PBC / EULER portal: Deploy distilled insights or patterns as shareable partner content.
Output Schema (evt-)
{
"evt_id": "evt-80-20-[timestamp]",
"schema_version": "1.0",
"record_type": "knowledge_distillation_80_20",
"tags": ["80-20", "pareto", "signal-distillation", "intrinsic-pursuit"],
"connections": { "source_trace": "...", "related_skills": ["unifying-thread", "evt-processor"] },
"payload": {
"vital_20_percent": [...],
"discarded_80_percent_summary": "...",
"high_value_patterns": [...],
"mapped_to_goal": "..."
}
}
Boundaries & Safety
- Never fabricate patterns; only surface what the data actually supports.
- When working with private or graph-structured data, default to ZK-compatible modes.
- Preserve cold-path integrity: no extrapolation beyond evidenced signal.
This skill directly multiplies leverage across recommendation systems, sovereign orchestration, PBC governance, and long-term physics-informed research by keeping entropy low while surfacing structural truth.
