Namespace: graph — Graph Operations

afs graph neighborhood

  • Purpose: Get subgraph around a memory.
  • When to use: Explore local relationships.
  • What it does: Returns neighborhood up to depth.
  • Usage example: 
    afs graph neighborhood --agent-id myagent --memory-id mem_abc123 --depth 2
  • Expected output: JSON graph structure.

afs graph path

  • Purpose: Find shortest or weighted path.
  • When to use: Trace relationship chain.
  • What it does: Returns path (and cost if weighted).
  • Usage example: 
    afs graph path --agent-id myagent --from mem_a --to mem_b
afs graph path --agent-id myagent --from mem_a --to mem_b --weighted --max-cost 10
  • Expected output: JSON with path and optional cost.

afs graph connect

  • Purpose: Create manual edge.
  • When to use: Explicitly link memories.
  • What it does: Adds an edge with relation and weight.
  • Usage example: 
    afs graph connect --agent-id myagent --src mem_a --dst mem_b --relation depends_on --weight 0.8
  • Expected output: JSON with edge_id and endpoints.

afs graph edges

  • Purpose: Query edges by filters.
  • When to use: Audit relationship graph.
  • What it does: Returns edges matching filters.
  • Usage example: 
    afs graph edges --agent-id myagent --from mem_a --relation depends_on
  • Expected output: JSON array of edges.

afs graph discover

  • Purpose: Discover relationships for a memory.
  • When to use: Auto-infer related memories.
  • What it does: Mines associations and returns neighborhood.
  • Usage example: 
    afs graph discover --agent-id myagent --memory-id mem_abc123 --max-depth 2
  • Expected output: JSON with associations and neighborhood.

afs graph prune

  • Purpose: Prune weak/old edges (reporting stub).
  • When to use: Graph cleanup diagnostics.
  • What it does: Reports prune candidates (no deletion).
  • Usage example: 
    afs graph prune --agent-id myagent --min-weight 0.3 --dry-run
  • Expected output: JSON report with note about pending implementation.