Slipway

Overview

How developers consume platform capabilities through a governed golden path.

Problem

Self-service delivery can fail without strong workflow controls. Slipway models deterministic planning, policy decisions, approval gates and execution evidence.

Engineering Goals

• Model a request-to-plan-to-approval lifecycle with reproducible state transitions.
• Keep policy decisions deterministic and server-side enforceable.
• Persist governance and execution evidence for auditability.

Architecture

• FastAPI backend with PostgreSQL persistence and migration-driven schema evolution.
• React frontend for request submission, plan review and controlled execution.
• Lifecycle entities for requests, plans, policy decisions, approvals, executions and audit events.

Key Engineering Decisions

• Separate planning, approval and execution concerns with explicit state checks.
• Use deterministic fingerprints and idempotency keys for replay-safe operations.
• Persist both successful and failed execution receipts for traceability.

Security / Reliability Controls

• Server-side policy and approval enforcement.
• Stale decision invalidation when inputs change.
• Execution immutability via API boundaries.

Validation / Testing

• Backend tests for policy, approval, execution and audit flows.
• Migration checks and frontend production build validation.
• CI-backed PostgreSQL lifecycle tests.

What This Demonstrates

• Internal developer platform workflow design.
• Governance with deterministic control-state logic.
• Developer experience balanced with operational controls.