Add sweep_workstream_prose.py for agent-guidance files, sweep active workplan prose in-repo, tighten scan allowlist exclusions, and update ADR-001 closure protocol to workplan-first terminology.
1.5 KiB
1.5 KiB
id: CUST-WP-0000b type: workplan title: "State Hub v0.2 — Decisions, Suggestions & Dependencies" domain: infotech repo: the-custodian status: completed owner: custodian topic_slug: custodian created: "2026-02-25" updated: "2026-02-25" completed: "2026-02-25" state_hub_workstream_id: "6585ee66-aa4e-436e-bbec-d83293c33e8f" note: > Pre-ADR-001 record. This workplan was created DB-first before the workplan-as-repository-artefact convention was established. Retroactive file written on 2026-03-11 to satisfy the ADR-001 consistency checker (C-08).---
State Hub v0.2 — Decisions, Suggestions & Dependencies
What was built
Delivered 2026-02-25, evolving the hub from a state tracker to an active coordination layer:
WorkstreamDependencymodel + migration0b547c153153— directed dependency graph between workplans- API:
POST/GET /workstreams/{id}/dependencies/,DELETE /workstreams/{id}/dependencies/{dep_id} - API:
GET /state/next_steps— derived next-action suggestions (never persisted) StateSummaryextended withnext_stepsanddepends_on/blockson workplans- Design boundary formalised: hub is a read model with exactly two write use cases — resolving decisions and suggesting next steps
- MCP:
get_next_steps()tool added scripts/script.py.makoadded (required for Alembic autogenerate)
References
- Alembic migration:
0b547c153153 - Design boundary document:
canon/architecture/(hub as read model) - CLAUDE.md global + railiance updated with
get_next_steps()in session start