write-be-spec

Overview

Runs BE specification authoring in two shards: classify → write.

Prerequisites

1. IA shard(s) are complete and marked ✅ in .memory/wiki/specs/ia/index.md
2. .memory/pipeline/progress/spec-pipeline.md exists
3. BE conventions/index files exist under .memory/wiki/specs/be/

Step-by-Step

Step 0 — Pipeline-state shard selection

1. Read .memory/pipeline/progress/spec-pipeline.md.
2. Find shards where IA is complete and BE is not-started.
3. Select lowest-numbered eligible shard unless user explicitly overrides.
4. If none remain, stop and recommend /write-fe-spec.

Step 1 — Run classify shard

Call skill: write-be-spec-classify

Expected outputs:

• Classification type + rationale
• Expected BE spec count
• Referenced Material Inventory
• Created BE spec stub(s) after approval

Step 2 — Run write shard

Call skill: write-be-spec-write

Expected outputs:

• Full BE spec file(s) in .memory/wiki/specs/be/
• Updated .memory/wiki/specs/be/index.md
• Updated .memory/pipeline/progress/spec-pipeline.md BE status
• Ambiguity gate results + next allowed command

Quality Gate

Do not close workflow unless all pass:

• request/response schema coverage per endpoint
• database schema/index/permission completeness
• middleware/security constraints reflected from IA
• specific error code definitions
• rate-limit specs per endpoint
• deep-dive and cross-reference coverage
• complete IA Source Map traceability

Completion Checklist

• target shard selected from pipeline state
• classify shard completed and approved
• write shard completed and approved
• BE index updated
• BE pipeline tracker updated
• ambiguity gate completed
• next valid command recommendation emitted

Next Steps

• if BE specs remain: /write-be-spec
• if BE complete and /audit-ambiguity be is clean: /write-fe-spec