Alpha MCPAPI reference · v1

The MCP your agents call.

Tritemporal memory that cannot lie about the past, a governed causal graph, knock-on simulation, and a decision ledger — the engine behind every Cascadian answer, at machine speed.

MCPPOSTGRESGOVERNED
Tools / Memory

bootstrap()

Key

The L0 read. One call at session start returns a ~200-token card: workspace state, open decisions, tripped invalidation criteria, and pending approvals. Everything else is ids to drill into — never a raw dump.

bootstrap()SESSION CARD

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

Why your agent calls it

The first call of every session — and usually the only one an agent needs to know where it stands. ~200 tokens back: which decisions are open, which invalidation criteria tripped while nobody was looking, what's waiting on a human. An agent that starts here never re-derives the workspace from scratch, and never misses a thesis that broke overnight.

What it returns
3OPEN DECISIONS1TRIPPED CRITERIA2PENDING APPROVALSL0 SESSION CARD · ~200 TOKENS · SUMMARIES + IDS, NEVER RAW DUMPS
bootstrap() · The ~200-token session card your agent reads first.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.bootstrap()
Parameters

None. The call is workspace-scoped by the key that makes it.

Example response
bootstrap()JSON · ILLUSTRATIVE
{ "workspace": "acme-revops", "open_decisions": [ { "id": "dec_0142", "q": "Cut the discount tier or raise onboarding spend?", "status": "awaiting_approval" } ], "tripped_invalidation_criteria": [ { "decision": "dec_0117", "criterion": "win_rate drop > 3pp in 60d" } ], "pending_approvals": 1, "tokens": 192 }
Related — Memory