Tasks: DRG completeness (#2843) — relation-description parity + activation-gate consolidation

Mission: drg-relation-parity-activation-gate-01KY48PD | Branch: doctrine/drg-completeness-2843 (coord topology) Spec: spec.md | Plan: plan.md | Research: research.md | Contracts: contracts/

Two parallel lanes: Item B (activation-gate live-bug fix — WP01→WP02/WP03) and Item A (relation-description parity — WP04→WP05). Item A ∥ Item B (disjoint files). Red-first (NFR-001) precedes the WP01 fix.

Subtask Index

IDDescriptionWPParallel
T001Red-first characterization test: stem-form RED + canonical-id GREEN control (real corpus + repo config)WP01[P]
T002Add drg.py-local singular→ArtifactKind constantWP01
T003filter_graph_by_activation: build resolved canonical-URN map once (lift _build_tension_active_urns); resolve_doctrine_root() + pack_roots[1:]; skip-with-report on UnknownArtifactIdErrorWP01
T004_node_is_activated: consume pre-resolved map, membership on full URN vs node.urnWP01
T005Root-source pinning test (gate root == compiler-projection root); green characterization; ruff/mypyWP01
T006Delete tension-scan trio; re-point tension consistency to consume the gate; verdicts unchangedWP02
T007Re-point _check_graph_kind_parity KIND→per-ID (behavior upgrade) + own testsWP02
T008Delete orphaned _DRG_SINGULAR_TO_ACTIVATED_KINDS_MEMBER; update dead-symbol/dead-module/arch baselinesWP02
T009executor.py:182 before/after test (named observable)WP03[P]
T010reference_resolver.py:67 before/after testWP03[P]
T011compiler.py:1037 closure before/after testWP03[P]
T012_check_drg_cross_kind_refs:424 before/after testWP03[P]
T013context.py:928 before/after testWP03[P]
T014Full tests/charter/ + tests/doctrine/ suite green — 0 net failures vs merge-baseWP03
T015Backfill RELATION_DESCRIPTIONS for the 12 members (applies≠scope adjudication; dormant vs overlay emission status)WP04[P]
T016Convert test_models.py =={3} → ==set(Relation) + non-empty over allWP04
T017Extend docs/context/doctrine.md glossary prose (non-enforced paraphrase)WP04
T018ruff/mypy/terminology/markdownlint on Item-A code + proseWP04
T019Restructure doctrine-relationships.md → 15 per-relation ### sections (7 new, split grouped enhances/overrides/replaces, trim Lineage/Delegation prose)WP05
T020Widen _SCOPED_RELATIONS 3→15WP05
T021Update stale "other twelve out of scope / follow-up" docstrings (test_relation_doc_parity.py + doc "Tension vocabulary" prose)WP05
T022Green test_relation_doc_parity.py (all 15 scoped) + terminology guardWP05

Work Packages

WP01 — Activation-gate canonical-URN correctness (IC-01) · Item B · Priority P1

  • Goal: Route the per-ID activation gate through the existing resolve_artifact_urn so a populated activated_directives (config stems) no longer silently drops canonical directive nodes. Red-first.
  • Independent test: uv run pytest tests/charter/test_drg_activation_gate.py -q — stem-form RED on merge-base, GREEN after; canonical-id control GREEN on merge-base.
  • Dependencies: none (Item B root). Prompt: tasks/WP01-activation-gate-canonical-urn.md
  • Subtasks: T001, T002, T003, T004, T005
  • □ T001 Red-first characterization test: stem-form RED + canonical-id GREEN control (real corpus + repo config) (WP01)
  • □ T002 Add drg.py-local singular→ArtifactKind constant (WP01)
  • □ T003 filter_graph_by_activation: build resolved canonical-URN map once; resolve_doctrine_root() + pack_roots[1:]; skip-with-report on UnknownArtifactIdError (WP01)
  • □ T004 _node_is_activated: consume pre-resolved map, membership on full URN vs node.urn (WP01)
  • □ T005 Root-source pinning test; green characterization; ruff/mypy (WP01)

WP02 — Workaround collapse (IC-02) · Item B · Priority P1

  • Goal: Delete the tension-scan reimplementation so it consumes the one gate; re-point _check_graph_kind_parity KIND→per-ID (intended behavior upgrade); delete the orphaned constant.
  • Independent test: tension-consistency verdicts unchanged; _check_graph_kind_parity per-ID tests green; dead-symbol/module + arch baselines green.
  • Dependencies: WP01. Prompt: tasks/WP02-workaround-collapse.md
  • Subtasks: T006, T007, T008
  • □ T006 Delete tension-scan trio; re-point tension consistency to consume the gate; verdicts unchanged (WP02)
  • □ T007 Re-point _check_graph_kind_parity KIND→per-ID (behavior upgrade) + own tests (WP02)
  • □ T008 Delete orphaned _DRG_SINGULAR_TO_ACTIVATED_KINDS_MEMBER; update dead-symbol/dead-module/arch baselines (WP02)

WP03 — Five-consumer regression net (IC-03) · Item B · Priority P1

  • Goal: Prove the corrected gate does not regress any of the five filter_graph_by_activation consumers, each with a named observable — None-path byte-identical, populated-path corrected.
  • Independent test: uv run pytest tests/charter/ tests/doctrine/ -q — 0 net failures vs merge-base.
  • Dependencies: WP01 (soft: WP02 for the _check_graph_kind_parity interaction — see risks). Prompt: tasks/WP03-consumer-regression-net.md
  • Subtasks: T009, T010, T011, T012, T013, T014
  • □ T009 executor.py:182 before/after test (named observable) (WP03)
  • □ T010 reference_resolver.py:67 before/after test (WP03)
  • □ T011 compiler.py:1037 closure before/after test (WP03)
  • □ T012 _check_drg_cross_kind_refs:424 before/after test (WP03)
  • □ T013 context.py:928 before/after test (WP03)
  • □ T014 Full tests/charter/ + tests/doctrine/ suite green — 0 net failures vs merge-base (WP03)

WP04 — Relation registry + completeness gate + glossary prose (IC-04 + IC-05b) · Item A · Priority P2

  • Goal: Complete RELATION_DESCRIPTIONS, convert the code-side completeness gate, and extend the non-enforced glossary.
  • Independent test: uv run pytest tests/doctrine/drg/test_models.py -q — ==set(Relation) + non-empty.
  • Dependencies: none (Item A root). Prompt: tasks/WP04-relation-registry-and-prose.md
  • Subtasks: T015, T016, T017, T018
  • □ T015 Backfill RELATION_DESCRIPTIONS for the 12 members (applies≠scope adjudication; dormant vs overlay emission status) (WP04)
  • □ T016 Convert test_models.py =={3} → ==set(Relation) + non-empty over all (WP04)
  • □ T017 Extend docs/context/doctrine.md glossary prose (non-enforced) (WP04)
  • □ T018 ruff/mypy/terminology/markdownlint on Item-A code + prose (WP04)

WP05 — Doc parity surface restructure (IC-05a) · Item A · Priority P2

  • Goal: Bring the single parity-enforced surface (doctrine-relationships.md) into lockstep and widen the parity test to all 15 relations.
  • Independent test: uv run pytest tests/doctrine/test_relation_doc_parity.py -q — green with _SCOPED_RELATIONS = 15.
  • Dependencies: WP04 (descriptions must exist first). Prompt: tasks/WP05-doc-parity-restructure.md
  • Subtasks: T019, T020, T021, T022
  • □ T019 Restructure doctrine-relationships.md → 15 per-relation sections (7 new, split grouped heading, trim prose) (WP05)
  • □ T020 Widen _SCOPED_RELATIONS 3→15 (WP05)
  • □ T021 Update stale non-goal docstrings (test_relation_doc_parity.py + doc "Tension vocabulary" prose) (WP05)
  • □ T022 Green test_relation_doc_parity.py (all 15 scoped) + terminology guard (WP05)

Dependencies & Lanes

Item B:  WP01 ──┬── WP02
                └── WP03
Item A:  WP04 ───── WP05
  • Item A ∥ Item B — disjoint files (Item A: doctrine/drg/models.py + docs/ + tests/doctrine/; Item B: charter/ + tests/charter/*), no overlap.
  • MVP: WP01 (closes the live correctness bug) is the highest-value slice; WP04 is the smallest independently-shippable Item-A slice.
  • Coord topology; finalize-tasks computes lanes.json from these dependencies.

Requirement coverage

  • Item B: FR-001, FR-002, FR-003, FR-004, NFR-001, NFR-002, C-001, C-002, C-003, C-004 → WP01/WP02/WP03
  • Item A: FR-005, FR-006, FR-007, FR-008, NFR-003, C-006 → WP04/WP05
  • NFR-004 (quality gates) → every WP.