AgentForce charges $2 per autonomous resolution. Every failed resolution is a retry, a CSAT hit, and $2 with no outcome. The billing trigger lives four hops from Einstein Agent — agents that don't traverse the full dependency chain get it wrong, and you pay for every wrong guess.

ckg-agentforce puts the full AgentForce stack into a traversable knowledge graph: 40 nodes, 52 edges, 4 typed relationship kinds (REQUIRES · ENABLES · IMPLEMENTS · RELATES_TO). Every edge was extracted from an official Salesforce source document and SHA-256 pinned to the exact bytes at extraction time. No probabilistic inference — declared edges only. If it's not in the graph, it returns nothing rather than a hallucinated approximation.

What it answers deterministically:

The exact prerequisite chain before AgentForce charges the $2 resolution fee
Which of the six capabilities gating on Einstein Trust Layer is blocking your agent
Why Grounding fails when Data Cloud isn't provisioned
The NVIDIA NIM integration path through Model Selection → Reasoning Engine

Benchmarked: F1 0.471 vs 0.123 RAG — 4× lift, 11× fewer tokens, $7.81 vs $76.23 per 1k queries. Billing path F1: +121%. Prereq-chain F1: +145%.

6 tools: list_concepts · search_concepts · query_ckg · get_prerequisites · resolution_path · verify_source

Free tier: 100 calls/day via this endpoint — no signup, no install, paste the URL directly into claude.ai or Cursor. When you hit the limit or need higher volume, the Dev plan ($29/mo) removes the cap. Pro ($99/mo) adds all 97 domains: NVIDIA, Finance, Healthcare, Regulatory, and Enterprise AI stacks.

Need a custom domain or enterprise deployment? Graphify.md builds CKGs for specific platforms, internal knowledge bases, and regulated stacks. Reach out at daniel.yarmoluk@gmail.com or graphifymd.com.