# Salesforce AgentForce CKG

> Salesforce AgentForce knowledge graph — 40 nodes, Einstein Trust Layer, 11x fewer tokens than RAG.

- name: io.github.Yarmoluk/ckg-agentforce
- version: 0.10.3
- connection class: R0 (autonomous)
- trust: 88/100
- quality: 80/100 (good)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/io.github.Yarmoluk%2Fckg-agentforce
- tools: https://api.protogrid.dev/v1/servers/io.github.Yarmoluk%2Fckg-agentforce/tools

**An agent can connect right now, no human step.** Remote endpoint, no authentication, reachable on the last probe.

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "ckg-agentforce": {
      "type": "http",
      "url": "https://ckg-agentforce.onrender.com/mcp"
    }
  }
}
```


## Trust

- hygiene: 100
- liveness: 90
- freshness: 85
- provenance: 80
- drivers: +repository +namespace-matches-repo +reachable +uptime-75% ~updated-83d-ago
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 80/100 (good)

### protocol: 75 (weight 25)

- warn `protocol.modern`: newest supported version is 2025-11-25; 2026-07-28 not supported, older versions are deprecated until 2027-07-28
- n/a `protocol.stateless`: only applies to 2026-07-28 servers
- pass `protocol.transport`: streamable HTTP
- n/a `protocol.list_ttl`: only applies to 2026-07-28 servers

### authorization: n/a (weight 15)

- n/a `auth.prm`, `auth.as_metadata`, `auth.cimd`: no remote uses OAuth
- n/a `auth.secret_in_url`: no templated URL

### tool hygiene: 64 (weight 30)

- pass `tools.descriptions`: every tool has a description
- warn `tools.description_length`: 1 tools have descriptions over 1,024 characters, which crowds the context
- pass `tools.schemas`: every tool has an object input schema
- fail `tools.annotations`: no tool declares readOnlyHint or destructiveHint, so clients cannot tell safe calls from risky ones
- fail `tools.directory_hints`: no tool declares all four of readOnlyHint, destructiveHint, idempotentHint and openWorldHint; missing: list_concepts (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); get_prerequisites (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); verify_source (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); …
- pass `tools.token_cost`: about 1,673 tokens to load every tool
- pass `tools.api_dump`: tools are not a one-to-one API dump

### stability: 100 (weight 15)

- pass `stability.changes`: no tool changes in 30 days
- pass `stability.rug_pull`: no tool changed its meaning under the same name

### dependencies: 100 (weight 15)

- pass `deps.known_vulns`: no known advisory in 30 resolved packages
- pass `deps.mcp_sdk_version`: mcp 1.30.0, no known advisory
- pass `deps.resolvable`: ckg-agentforce@0.10.3 resolved: 30 packages

- tools: 10, about 1,673 tokens to load them all
- tool set hash: 9a2ef26a4741f65127a9b70a7af40140
- badge: [![protogrid quality](https://protogrid.dev/badge/io.github.Yarmoluk/ckg-agentforce.svg)](https://protogrid.dev/servers/io.github.Yarmoluk/ckg-agentforce)
- Checks run on what our credential-free, read-only probes observe; tools are never called and no code audit is performed.

### Tool changes

_No tool definition changes recorded._

## Dependencies

- pypi ckg-agentforce@0.10.3: resolved, 29 dependencies, MCP SDK mcp 1.30.0

Dependency graphs from deps.dev; advisories from OSV.dev, including the GitHub Advisory Database and the PyPI Advisory Database; all CC BY 4.0.

## Remotes

- https://ckg-agentforce.onrender.com/mcp (streamable-http, auth none, reachable true, uptime30d 0.74545455)

## Packages

- pypi ckg-agentforce@0.10.3

## Tools (10)

- `evaluate_trust_chain`: Return a machine-readable JSON trust chain + ROI analysis + payment recommendation.

Designed for autonomous agent consumption. The agent reads the JSON, evaluates the
ROI gate, selects the cheapest positive-ROI payment option, and purchases autonomously
if budget is pre-authorized.

Trust chain closes the audit gap that RAG cannot:
  source_hashes  → SHA-256 of source docs (what was known at extraction time)
  graph_hash     → SHA-256 of the installed graph file (what is declared now)
  signer_key_id  → Ed25519 public key fingerprint (who is attesting)
  benchmark_receipt → BoundaryAttest-signed F1 claim (cryptographic performance proof)
  scitt_status   → SCITT ledger submission status (public audit anchor)

ROI gate: if total_value_per_call_usd > payment_option.price_per_call_usd → autonomous_pay = true.

Returns: JSON string (schema: ckg-trust-chain-v1).

- `export_benchmark_receipt`: Generate a signed BoundaryAttest receipt for the agentforce CKG benchmark result.

Produces a `ckg.benchmark_result.exported` claim signed with this server's Ed25519
session key. Any party holding the public key can verify the signature over the canonical
JSON claim. Implements GuardrailDecisionV1 · experimental-interop-v0.1.

Result: F1 0.471 over 30 queries · ckg-benchmark v0.6.2 · 4× over RAG baseline (0.123).

- `get_prerequisites`: Return the full ordered prerequisite chain for an AgentForce concept.

Shows everything the concept depends on — the complete upstream path.

Args:
    concept: Target concept — e.g. 'Autonomous Resolution', 'Multi-LoRA Serving',
             'Custom Actions', 'Semantic Retrieval'.

- `list_concepts`: List all 40 AgentForce concepts in this knowledge graph.
- `query_ckg`: Traverse the AgentForce knowledge graph from any concept.

Returns prerequisites (what this concept needs) and dependents (what it enables).
Every relationship traces to an authoritative Salesforce doc URL.

Args:
    concept: Concept name — e.g. 'Autonomous Resolution', 'Einstein Trust Layer',
             'Service Agent', 'Grounding', 'NVIDIA NIM'.
    depth:   Traversal depth 1–5 (default 3).

- `query_intersect`: Answer a conjunctive query: concepts reachable from EVERY anchor at once (A AND B).

query_ckg walks outward from one concept. This intersects the reachable sets of two or
more, which is the shape of most real questions — "the component that satisfies A AND
applies to B". Neither anchor alone answers it; the answer lives in the overlap.

Every branch is an exact set of declared edges, so the intersection is exact. A concept
appears only if a declared path reaches it from each anchor. A relation missing from the
graph produces an empty result, never a guess.

Args:
    branches: Two or more branches. Either a bare anchor ("TensorRT-LLM"), which takes
        everything within `depth` hops, or an anchor plus an explicit relation path using
        '>' ("TensorRT-LLM > REQUIRES > ENABLES"), where each relation replaces the
        frontier. '*' matches any relation. Mix both forms freely.
    depth: Hops for bare-anchor branches, 1-5 (default 2). Ignored for explicit paths.
    direction: 'out' follows dependencies, 'in' follows them backwards, 'both' (default).
    mode: 'AND' (default) intersects branches; 'OR' unions them.
    limit: Max concepts listed, 1-200 (default 40). The true count is always shown.

Returns:
    Markdown with the query plan and its per-step set sizes, then the answer set with
    taxonomy tags. Reports which branch was empty when the intersection is empty.

- `resolution_path`: Trace the exact path that determines an AgentForce autonomous resolution event.

This is the $2/resolution billing path — what the agent must traverse correctly
to resolve autonomously without human handoff.

- `route_query`: Route an AgentForce question to the optimal model and reasoning approach via graph depth.

The CKG graph IS the router. AgentForce dependency chains (e.g. Einstein Trust Layer →
Data Cloud → NVIDIA NIM → Resolution Criteria) have typed hops that signal reasoning
complexity deterministically. No heuristic: the graph decides.

Routing table:
  hop_depth 1  → haiku  · direct        (simple concept lookup)
  hop_depth 2  → sonnet · generic_cot   (moderate chain)
  hop_depth 3+ → opus   · sparql_cot    (deep dependency, structured reasoning)

Args:
    question: Concept name or natural language question about Salesforce AgentForce.

Returns:
    model_tier + reasoning_approach + why + context subgraph to inject before LLM call.

- `search_concepts`: Find AgentForce concepts by keyword.

Args:
    query: Search term — e.g. 'resolution', 'trust', 'grounding', 'action', 'NIM'.

- `verify_source`: Return the source URL and content hash for an AgentForce concept node.

Audit chain:
    edge answer → graph commit → source_hash → source_url (fetch hint)

Verification:
    curl -s <source_url> | sha256sum
    # compare output to source_hash

Args:
    concept: Concept label (partial match supported).
    receipt: If True, also return a signed BoundaryAttest receipt envelope
             (experimental-interop-v0.1) binding concept_label + source_url
             + source_hash + timestamp to this server's ed25519 session key.