Current direction and active tensions. Historical ship data is in CHANGELOG.md.
- Implementing the "Weaver" phase: task dependency graphs, frontiers, and critical path analysis.
- Modeling stories, requirements, acceptance criteria, and evidence as native graph nodes.
- Auto-detecting test-to-requirement links through the scanning pipeline.
- Hardening the agent-native CLI (
briefing,next,context,act). - Implementation of the versioned JSONL control plane for speculative worldline management.
- Integration of the MCP (Model Context Protocol) server for native agent participation.
- Refinement of the Sovereignty Audit to verify the complete Genealogy of Intent.
- Formalization of the "Oracle" phase: intent classification and policy engine enforcement.
- Schema Rigidity: Adjusting to the evolving Digital Guild vocabulary while maintaining backward compatibility for existing WARP patches.
- TUI Density: Balancing information richness in the Bijou cockpit with the "trapdoor" capture doctrine.
- Speculative Complexity: Managing the cognitive load of derived worldlines and braiding operations for agents.
- Environment Parity: Ensuring bit-identical behavior between the standalone CLI actuator and the JSONL API.
With CQRS Intent-Based UI Architecture (v1.0.0-alpha.16) fully complete,
the immediate #1 priority is Substrate Decoupling & Bedrock Purity -
asserting the absolute architectural law that XYPH MUST NOT DO GIT-WARP'S
JOB. Xyph should only know about Optics and Intents. Worldline
forking/braiding is permissible, but Xyph must NEVER imperatively manage
nodes, graphs, traversals, or materialization state machines.
The current design topic is GraphQL Optics, Intent Declarations, and the Edict Bridge: domain GraphQL plus XYPH lawpacks compiled by Wesley now, with Edict as the future execution/law substrate. The active kill list is Substrate Decoupling Boundary Audit.