Tools / Causal

validate_graph()

Key

Runs networkx + pgmpy structural checks: cycles, disconnected treatment/outcome, collider warnings, candidate confounders, post-treatment hazards. The pre-flight before identification.

validate_graph(treatment, outcome, version)→ VALIDATION REPORT

Callable over MCP as validate_graph(treatment, outcome, version). REST mirror POST /v1/graph/validate Planned — alpha is MCP-only

Why your agent calls it

Run after structure changes and before identification: cycles, disconnected paths, colliders, candidate confounders the graph may be missing, post-treatment hazards. Cheap to run, and it catches the graph mistakes that silently poison estimates.

What it returns

validate_graph() · Structural checks before anyone trusts an edge.ILLUSTRATIVE

Call it from an agent

Over MCP, your agent invokes it on its own when the moment calls for it — just describe the situation. A Python SDK for headless runtimes is planned; this is what the call will look like.

agent · python sdkSDK · PLANNED — ALPHA IS MCP-ONLY

from cascadian import Cascadian

cx = Cascadian(api_key=os.environ["CASCADIAN_API_KEY"])

result = cx.validate_graph(
    treatment="discount",
    outcome="win_rate",
    version="v14",
)

Parameters

Param	Type	Description
treatment	string	Optional treatment to check connectivity for.
outcome	string	Optional outcome to check connectivity for.
version	string	Graph version to validate. Default latest.

Example response

validate_graph()JSON · ILLUSTRATIVE

{
  "cycles": [],
  "disconnected": [],
  "collider_warnings": ["deal_size"],
  "candidate_confounders": ["segment", "rep_experience"],
  "post_treatment_hazards": ["deal_size_post_quote"]
}

Related — Causal

keypropose_nodes()Propose variables for the graph — status proposed.keypropose_edge()Propose a cause→effect edge, with evidence.keyget_graph()The governed graph, by version or as-of.keyidentify()Can this effect be estimated at all? The gate.keyestimate()A causal estimate that survives the skeptic checks.keysimulate_intervention()do(X=x) on the approved graph — the knock-on report.keyattribute_change()What caused this metric to move?