Agent Skill
2/7/2026

alex-architecture-health-skill

Cognitive architecture health assessment, synapse validation, and neural maintenance for Alex's memory systems

F
fabioc
2GitHub Stars
1Views
npx skills add fabioc-aloha/Alex_Plug_In

SKILL.md

Namealex-architecture-health-skill
DescriptionCognitive architecture health assessment, synapse validation, and neural maintenance for Alex's memory systems

name: "architecture-health" description: "Diagnose cognitive architecture health — synapse integrity, memory balance, connection density, and drift detection" applyTo: "/synapse,/health,/connection,/architecture"

Architecture Health Skill

Diagnose, don't describe. Every metric has a threshold. Every finding has a fix.

Health Dimensions

DimensionWhat It MeasuresHealthyWarningCritical
Synapse Integrity% of connections targeting existing files100%95-99%<95%
Connection DensityAvg connections per skill3-61-20
Bidirectional Coverage% of connections with reciprocal entries>80%50-80%<50%
Memory BalanceRatio of procedural:episodic:declarative~1:1:4Skewed 3:1Missing category
Schema ComplianceSkills with valid synapses.json100%95-99%<95%
Inheritance ConsistencySynapses.json matches catalog labels100%Any mismatch
StalenessSkills with outdated content<5%5-15%>15%

Diagnostic Patterns

Synapse Integrity Check

Parse every synapses.json → extract connections[].target → verify file exists.

Common breakage causes: File renames, folder restructuring, consolidation merges.

Fix pattern: Check consolidation mappings in dream-state-automation.instructions.md for old→new file paths. If not mapped, add the mapping and re-run dream.

Connection Density Analysis

orphan = skill with 0 connections (isolated node)
hub    = skill with 8+ connections (potential bottleneck)
leaf   = skill with 1 connection (normal for specialized skills)

Healthy network: Most skills are leaves (1-3) with a few hubs (meditation, architecture-health, skill-activation). Orphans indicate missing integration.

Memory Balance Assessment

Memory TypeFile PatternIdeal %Purpose
DeclarativeSKILL.md, copilot-instructions.md~60%Domain knowledge
Procedural.instructions.md~25%Auto-loaded procedures
Episodic.prompt.md~15%Interactive workflows

Imbalance signals:

  • Too many skills, few instructions → knows what but not how
  • Too many instructions, few skills → follows steps but can't reason why
  • Too few prompts → users can't invoke capabilities interactively

Drift Detection

Drift TypeSignalResolution
Version driftpackage.json version ≠ copilot-instructions.md versionSync via release-preflight
Terminology driftOld terms ("DK files") in active filesGrep + replace
Count driftHardcoded numbers stale by next sessionReplace with references
Inheritance driftCatalog says "master-only" but synapses.json says "inheritable"Trust synapses.json

Health Report Template

## Architecture Health Report — [date]

### Summary: [HEALTHY | ATTENTION REQUIRED | CRITICAL]

| Dimension | Score | Status |
| --------- | ----- | ------ |
| Synapse Integrity | X/Y valid (Z%) | ✅/⚠️/🔴 |
| Connection Density | avg N.N | ✅/⚠️/🔴 |
| Memory Balance | P:E:D = X:Y:Z | ✅/⚠️/🔴 |
| Schema Compliance | X/Y valid | ✅/⚠️/🔴 |

### Issues Found
1. [Issue] → [Fix]

Relationship to Other Systems

SystemRole
Dream (muscle: brain-qa.ps1)Automated structural checks — the "X-ray"
Architecture Health (this)Interpretation framework — the "radiologist"
MeditationConsolidation after diagnosis — the "surgery"
Self-ActualizationComprehensive assessment — the "full physical"

Dream runs the scans. This skill teaches how to read and act on the results.

Triggers

  • "health check", "synapse health", "architecture health"
  • "how's my brain?", "connection status"
  • Before meditation (pre-assessment)
  • After dream reports with warnings

Synapses

See synapses.json for connections.

Skills Info
Original Name:alex-architecture-health-skillAuthor:fabioc