get_graph_digest
Pre-computed graph analytics in one call: counts, health, chain completeness, business-area coverage, lifecycle balance. ~500 tokens vs ~5-8K for equivalent manual fetches.
Context & Sessionatomic (read-only)
Arguments
if_changed_sincestringoptionalHash from a previous response. Returns { changed: false } if graph unchanged (saves ~470 tokens).
Returns
Shape
{ counts, health, chains, coverage, lifecycle, lens, lens_digest, _hash }~500 tokens vs ~5-8K for equivalent manual fetches.
Examples
Live call against the Notion example graph.
Output
{
"product": {
"title": "Notion (SATURATED test graph)",
"stage": "concept"
},
"counts": {
"total_nodes": 2054,
"total_edges": 3216,
"by_type": {
"product": 13,
"capability": 20,
"value_stream": 3,
"decision": 13,
"evidence": 10,
"assumption": 30,
"outcome": 6,
"document": 9,
"objective": 6,
"key_result": 14,
"design_component": 15,
"hypothesis": 6,
"experiment": 5,
"experiment_plan": 13,
"job": 10,
"bounded_context": 13,
"initiative": 11,
"strategic_pillar": 6,
"test_plan": 8,
"screen": 15,
"experiment_run": 18,
"design_system": 4,
"team": 1,
"compliance_requirement": 5,
"workspace": 3,
"epic": 4,
"bug": 6,
"task": 3,
"feature_area": 3,
"mission": 2,
"desired_outcome": 4,
"model_comparison": 3,
"agent_session": 3,
"learning": 17,
"research_plan": 6,
"constraint": 3,
"stakeholder": 10,
"feature": 12,
"security_review": 8,
"metric": 23,
"person": 215,
"technical_debt_item": 12,
"risk": 9,
"service": 11,
"opportunity": 7,
"solution": 11,
"feasibility_study": 3,
"design_sprint": 3,
"prototype": 8,
"design_concept": 7,
"vision": 1,
… (truncated)