Issue-driven control plane for supervised agent-assisted software development, with planner manifests, isolated worktrees, lifecycle gates, CI/review checks, and guarded merge flows.

A Python stdlib-only scaffold that turns AI-assisted development into verified engineering: structured task briefs, evidence bundles, risk levels L0–L4, and active quality gates (interface contracts, test quality, time allocation). Conversation discovers intent; artifacts control implementation; evidence controls acceptance.

Formal verification framework for dataset lineage, policy compliance, and training-time safety guarantees.

SCPA Governance Center: a deployable control plane that provides governed, scoped service context to AI agents through MCP, REST APIs, portal UI, schema validation, CI gates, and OPA policies.