Tools / Governance

reconstruct_decision()

Key

Re-runs the decision's pinned as-of queries today and diffs the results against the pinned ids. An empty diff proves the context the decision saw is exactly reconstructible; a non-empty diff is the audit artifact.

reconstruct_decision(decision_id)→ AS-OF DIFF

Callable over MCP as reconstruct_decision(decision_id). REST mirror POST /v1/reconstruct Planned — alpha is MCP-only

Why your agent calls it

The audit, executable. Re-run the exact pinned reads from decision time and diff against what was pinned. An empty diff is proof the record is faithful; a non-empty one shows precisely what changed underneath. "What did we know when we decided?" stops being a debate.

What it returns

reconstruct_decision() · Re-run the pinned reads. The diff is the audit.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.reconstruct_decision(
    decision_id="dec_0142",
)

Parameters

Param	Type	Description
decision_idREQUIRED	string	The decision to reconstruct.

Example response

reconstruct_decision()JSON · ILLUSTRATIVE

{
  "decision_id": "dec_0142",
  "reran": 14, "matched": 14, "diff": [],
  "verdict": "the context the decision saw is exactly reconstructible"
}

Related — Governance

scopedrequest_approval()Ask the human. The token never crosses this boundary.keyget_pending_approvals()What is waiting on a human.keydetect_leakage()Catch any future fact that leaked into the past.publiccapabilities()The honesty endpoint — what is real vs planned.