Checklists

requirements.md

Specification Quality Checklist: Meta.json Fail-Closed Read Routing

Purpose: Validate specification completeness and quality before proceeding to planning Created: 2026-08-10 Feature: spec.md

Content Quality

  • ✅ No implementation details (languages, frameworks, APIs) — inherent-domain note: this is an internal tech-debt mission whose subject IS the meta.json read seam; requirements name behavioral outcomes (fail loud, one verdict, honest gates), and the seam/gate references in Constraints are the mission's domain boundary, not premature implementation choices.
  • ✅ Focused on user value and business needs (fail-loud on corruption; no contradictory lock verdicts; trustworthy gates)
  • ✅ Written for the affected stakeholders (maintainers and agents on ref-advance / implement / merge paths)
  • ✅ All mandatory sections completed

Requirement Completeness

  • ✅ No [NEEDS CLARIFICATION] markers remain (both operator decisions resolved: C-005 absent≠null, C-006 deviation record)
  • ✅ Requirements are testable and unambiguous
  • ✅ Requirement types are separated (Functional / Non-Functional / Constraints)
  • ✅ IDs are unique across FR-###, NFR-###, and C-### entries
  • ✅ All requirement rows include a non-empty Status value
  • ✅ Non-functional requirements include measurable thresholds (symbol counts, 0 silent absorptions, gates green from live measurement)
  • ✅ Success criteria are measurable
  • ✅ Success criteria are technology-agnostic where possible (counts of routed sites, silent absorptions, decoder/comparator symbols, gate state)
  • ✅ All acceptance scenarios are defined
  • ✅ Edge cases are identified
  • ✅ Scope is clearly bounded (5 sites, one comparator, one governance record; ordered #3229→#3230; #3228 + #3240 folded in)
  • ✅ Dependencies and assumptions identified (C-001 ordered dependency, C-007 branch-reality reconciliation)

Feature Readiness

  • ✅ All functional requirements have clear acceptance criteria
  • ✅ User scenarios cover primary flows (P1 fail-loud, P2 unified comparator, P3 honest gates)
  • ✅ Feature meets measurable outcomes defined in Success Criteria
  • ✅ No implementation details leak into specification beyond the inherent-domain seam/gate boundaries noted above

Notes

  • Items marked incomplete require spec updates before /spec-kitty.plan.
  • All items pass. The two operator decisions were resolved during discovery (C-005: distinguish absent from null; C-006: record the #3240 deviation rather than register a new baseline).
  • Post-spec adversarial squad (2026-08-10, 4 lenses: structure / anti-laziness / live-gate / sequencing) folded in. Convergent corrections added: kernel placement of L1 + comparator (C-008, so git-plumbing site A can route without importing specify_cli); ROUTED_CALLEES extension + margin-based floor re-derivation (C-002/FR-008); malformed-vs-empty boundary (C-010); per-module atomic WPs + floor serialization (C-001/C-009); captured red-first with typed error + site identifier (FR-007); enumeration + completeness gate (FR-010/NFR-001); inline-literal field-set hole closed (NFR-002); ref_advance-import ratchet (NFR-004); corrected 5-site taxonomy; FR→site / issue-closure map. Squad confirmed the mission's premises are sound (3 gates green; site A unrouted; diagnosability tests absent).