analyse

Analyse – Knowledge Base Status

Read-only inspection of the current knowledge base. Reports what is known, what is missing, and suggests next actions. This skill never writes files.

Invocation

Invocation	Action
/analyse or /analyze	Full status report
/analyse gaps	Focus on gap matrix only
/analyse graph	Focus on knowledge graph statistics only

Procedure

1. Check if synthesis data exists

• Look for synthesis/graph/nodes.yaml and synthesis/graph/edges.yaml
• If not found: Report that no research has been conducted yet, suggest /research TOPIC to start

2. Knowledge Graph Statistics

Read synthesis/graph/nodes.yaml and synthesis/graph/edges.yaml:

• Node counts: Total nodes, by type (strategy, mechanism, constraint, risk, role, structure)
• Edge counts: Total edges, by type (supports, conflicts, requires, amplifies, activates, limits, backs, rebuts)
• Connectivity: Isolated nodes (0 edges), bridge nodes (connect clusters), most-connected nodes (top 5)
• Edge density: edges / (nodes × (nodes-1) / 2) — how interconnected is the graph?

3. Argumentation Health

If any nodes have qualifier set (i.e., they are tracked claims):

• Claims by qualifier level: Count distribution across gesichert / hoch / mittel / niedrig / spekulativ
• Unsupported claims (red flag): Claims with qualifier set but no backs edges pointing to them — no formal evidence trail
• Claims without rebuttal (yellow flag): Claims with qualifier set but rebuttal is null — untested assumptions
• Untested claims (actionable): Claims with validates_with set but qualifier still niedrig or spekulativ — we know how to test but haven't
• Edges without warrant: Count of edges (especially backs and rebuts) where warrant is null — unexplained relationships
• Stale claims: Claims whose qualifier has not changed despite dependent data (backing/rebutting nodes) being updated

If no nodes have qualifier set, skip this section silently.

4. Argument Chain Health

If synthesis/graph/chains.yaml exists:

• Chains by provenance: Count of literature / synthesis / hypothesis chains
• Composite qualifier distribution: Count across gesichert / hoch / mittel / niedrig / spekulativ
• Weakest links: List the weakest links across all chains — these are the top research priorities
• Chains without alternatives (single-explanation risk): Chains where alternatives is empty
• Cross-disciplinary innovation score: Chains crossing 3+ disciplines (most novel insights)
• Alternative chain comparison: Where alternatives exist, show relative qualifier ranking

If no chains exist, skip this section silently.

5. Morphological Box Coverage

Read synthesis/zwicky/dimensions.yaml and synthesis/zwicky/matrix.md:

• Dimensions: List all dimensions with option counts
• Coverage: Filled cells vs. unfilled cells (percentage)
• Sparse dimensions: Dimensions with fewer than 2 options (need research)

6. Gap Analysis

Read synthesis/gaps/gap-matrix.md and synthesis/gaps/research-questions.md (if exists):

• Gap counts: HIGH / MEDIUM / LOW priority gaps
• Top 5 gaps: Show the highest-priority gaps with their research questions
• Open research questions: List unresolved questions from previous runs

7. Research History

Read synthesis/runs/ directory:

• Run timeline: List all runs with date, type, and key insights
• Last run: When was the last research conducted?
• Trend: Is knowledge growing? (compare node/edge counts across runs if delta files exist)

8. Suggested Next Actions

Based on the analysis, suggest concrete next steps:

• If many HIGH gaps exist: "Consider /research TOPIC for: X, Y, Z"
• If graph is sparse: "The graph has few connections. Consider researching bridging topics."
• If coverage is high: "Knowledge base is well-covered. Consider /synthesis to derive strategies."
• If no synthesis has been done: "Consider /synthesis to generate combinations and decision trees."
• If unsupported claims exist: "Consider adding backs edges to formalize evidence trails for: X, Y, Z"
• If chains have validated weakest links: "Consider /assess to check if chain confidence can be upgraded"
• If many claims lack rebuttals: "Consider defining rebuttal conditions — what would invalidate these claims?"

Output Format

Structure the report clearly with sections. Use tables for statistics. Example:

## Knowledge Base Status

### Graph: 24 nodes, 31 edges
| Type | Count |
|---|---|
| strategy | 8 |
| mechanism | 6 |
| ... | ... |

Edge density: 0.11 (sparse — many potential connections unexplored)
Isolated nodes: node-x, node-y

### Coverage: 67% (18/27 cells filled)
Sparse dimensions: "Regulatory Framework" (1 option)

### Top Gaps (3 HIGH, 5 MEDIUM, 2 LOW)
1. [HIGH] Regulatory compliance in EU markets → "What are the key regulatory barriers?"
2. [HIGH] Supply chain resilience → "How do different suppliers compare on reliability?"
3. ...

### Last Research: 2026-03-20 (depth: supply-chain)
Trend: +4 nodes, +7 edges since previous run

### Suggested Next Steps
- `/research regulatory-compliance` — highest-priority gap
- `/synthesis` — enough data to generate initial combinations

Important

• This skill is read-only — it never creates or modifies files
• All output in English
• Distinguish between facts (from data) and suggestions (your recommendations)
• If synthesis data is incomplete or missing, say so clearly rather than guessing