3.2.x Goal Corroboration — Planning / Tracking POV (planner-priti)

Author: Planner Priti (work-decomposition & delivery-sequencing specialist). Lens: git-meta + tracker record (missions, PRs, issues, commit counts), NOT code-design. Range: v3.1.10 (6975ee2, 2026-06-04) .. v3.2.0 (40e5209, 2026-06-16) — 2317 raw commits. Claim under test: the operator's 3.2.x goals (G1 doctrine→runtime depth; G2 core-domain strangler onto SSOTs; G3 DevEx & enablers, per docs/release-goals/3.2.x.md) are evidence-grounded continuations of a 3.2.0 trajectory, not net-new pivots.

Directives Applied (planner-priti)

  • Directive 003 — Decision Documentation Requirement: every goal→evidence mapping and every verdict below carries its cited git/tracker rationale (commit ranges, path counts, mission slugs, PR/issue numbers). No verdict is asserted without a traceable count.
  • Modes engaged: decomposition (the in-range mission inventory broken to goal buckets), sequencing (trajectory-into-tag analysis), prioritisation (strongest/weakest evidence ranking), risk-analysis (where evidence is thin / classification is soft).
  • Avoidance boundary honoured: I do NOT adjudicate code-design correctness of the SSOTs (that is architect-alphonso / randy-reducer territory — see naming-identity-ssot-strangler/). I corroborate the delivery record only.

Method & Caveats (so the numbers are honest)

  • Mission inventory = kitty-specs/*/spec.md files added in-range (git log --diff-filter=A): 129 missions landed (excludes test-feature-* E2E leak dirs + 1 empty stub).
  • Goal classification of missions is slug-keyword-based with a fixed precedence (G2-strangler > G1-doctrine > G3-devex > other/enabler). This is a planning heuristic, not a semantic read of each spec.md; soft edges are flagged in Risks. Counts are directional, not exact.
  • Commit counts are path-scoped to real src/specify_cli/<domain> modules (agent-dir copies .claude/.codex/.amazonq/…, vendored, locks excluded per scope). op(<profile>): … governance commits (43) are workflow audit-trail, not feature delta.
  • PRs: 119 merged in merged:2026-06-04..2026-06-17 (gh, authoritative). ~63 of those are release/rc/tranche plumbing (3.2.0aN..rc42..3.2.0) — i.e. the entire range is the 3.2.0 push, which is itself the corroboration that the goals describe the 3.2.0 line of work.
  • Authorship: 1858/2317 commits Robert Douglass, 421 Stijn Dejongh — a concentrated, intentional release drive, not scattered drive-by.

Goal × Evidence Table

Goal In-range missions (slug-classified) Path-scoped commit signal (src/specify_cli/…) Merged-PR signal Guard/ratchet test surface Verdict
G1 Doctrine/Charter/DRG → runtime 26 (charter ×9 tranche/p7/pack-activation, doctrine/glossary ×8, DRG ×3, profile ×3, spdd/reasons ×1, lifecycle-dispatch-DRG-closeout) doctrine+charter+drg paths 157; mission_step_contracts 17; charter_runtime 8; charter_activate 6; doctrine-test suite 204 files ~14 (charter/doctrine/glossary/profile-projection) mission_step_contracts/test_executor.py, DRG node-kind + step-contract-runtime-input fixes (#1518/#1523) SUPPORTED (depth qualifier)
G2 Core-domain strangler → SSOT 28 (execution-state ×3, coordination/topology ×3, merge ×4, status/lane ×5, identity/naming ×4, ownership ×3, canonical-producer ×2, event-arch, mission-state-migration) status 77, core 61, missions 58, lanes 35, coordination 42, merge 14, mission_runtime(src/) 8, identity 4, context 3 21 test_execution_context_parity.py, test_merge_pipeline_ratchets.py, test_ratchet_baselines.py, test_shared_package_boundary.py — all extended in-range SUPPORTED (strongest)
G3 DevEx & enablers 19 (test-suite-acceleration, mutant-slaying, complexity/code-smell, tool-surface-contract ×2, tooling-stability-guard-coherence, stabilization/blocker-cleanup ×4, p0/p1 fixes) tests/** 760 (architectural ratchets 81), .github/workflows/** 72 37 (releng/sonar/ci/lint/hooks/parallel) HOW_TO_MAINTAIN.md + docs/release-goals/ convention landed; test_no_legacy_terminology.py, parity ratchets in active use SUPPORTED

Mission Inventory Mapped to Goals (in-range, 129 total)

G2 — Strangler / SSOT (28) — the spine of the release: execution-state-domain-remediation-01KT6HVH · execution-state-canonical-surface-01KTG6P9 · execution-context-unification-01KTPKST · coordination-merge-stabilization-01KTXRVR · coordination-topology-stabilization-01KTZVQ2 · mission-coordination-branch-atomic-event-log-01KSPTVW · wp-lane-state-machine-fsm-01KTGZAZ · status-writepath-profile-surface-remediation-01KTB6AN · mission-state-migration-determinism-cleanup-01KRC7JG · name-vs-authority-remediation-01KTYGTE · mission-identity-seam-and-1908-panel-01KV6510 · identity-boundary-ci-gate-rerun-01KS51YK · unblock-sync-identity-boundary-canary-01KRZJ07 · functional-ownership-map-01KPDY72 · charter-ownership-consolidation-and-neutrality-hardening-01KPD880 · migration-shim-ownership-rules-01KPDYDW · canonical-producer-lint-01KS4XX4 · canonical-producer-refactor-01KS7PEK · merge-done-surface-resolver-01KTDVHZ · merge-preflight-remote-state-boundary-separation-01KTBE5M · merge-review-status-hardening-sprint-01KQFF35 · review-merge-gate-hardening-3-2-x-01KRC57C · event-architecture-cli-git-truth-01KT119Y · backward-transition-cli-emit-01KRV8GC · status-commit-rename-01KSPN6C · sync-diagnose-canonical-allowlist-01KS4F8H · runtime-mission-execution-extraction-01KPDYGW · locate-project-root-consolidation-01KV5SX1.

G1 — Doctrine/Charter/DRG (26): charter-* (828-sprint, contract-cleanup, e2e-followups, end-user-docs, golden-path-e2e, mediated-doctrine-selection, p7-release-closure, p7-schema-versioning, pack-activation-layer, ux-and-org-pack-vocabulary) · doctrine-enrichment-frontend-brownfield-normalization · doctrine-glossary-architecture-consolidation-01KTNWFC · glossary-drg-chokepoint · glossary-drg-surfaces-and-charter-lint · glossary-functional-module-extraction · glossary-seed-file-schema-validation · layered-doctrine-org-layer-01KRNPEE · mission-lifecycle-dispatch-drg-closeout-01KV0S99 · org-doctrine-profile-integrity-{activation-closure,closeout} · phase-3-charter-synthesizer-pipeline · profile-invocation-runtime-audit-trail · profile-roles-as-value-object · spdd-reasons-doctrine-pack · agent-profile-projection-plugin-production · pre-doctrine-test-stabilization.

G3 — DevEx / enablers (19): test-suite-acceleration-01KV3H59 · mutant-slaying-core-packages · complexity-code-smell-remediation · p0-test-failure-resolution-1298-1305 · p1-dependency-cycle-cleanup · phase6-composition-stabilization · quality-devex-hardening-3-2-01KRJGKH · stability-and-hygiene-hardening-2026-04 · stabilization-release-core-bug-fixes · stable-320-{p0-cli-stabilization,release-blocker-cleanup} · tool-surface-contract-01KV2K2P · tool-surface-contract-residuals-01KV4S5B · tooling-stability-guard-coherence-01KTRC04 · test-stabilization-{and-debt-pass,pre-existing-cluster-fix} · sync-rejection-classification-and-queue-retry-hygiene · task-workflow-bug-fixes · cli-bug-sweep-tool-surface-self-registration · private-teamspace-ingress-safeguards.

other/enabler (~56): auth/sync/teamspace/SaaS hardening, docs/install/upgrade UX, mission-composition rewrites, session-presence, pi-and-letta agent support, codebase-sanitization, do-dispatch lifecycle, release tranches. These are non-goal-or-adjacent delivery (the operator's own §Non-goals defers UX/SaaS to 3.3.x); their volume is the noise floor against which the three goals stand out, not counter-evidence.

Per-Goal Corroboration Verdict

G1 — SUPPORTED, with a depth qualifier

Quantitatively undeniable as a direction: 26 in-range missions and 157 doctrine/charter/DRG path-commits, plus a 204-file doctrine test corpus and concrete runtime-wiring commits (mission_step_contracts/test_executor.py; step-contract-runtime-input #1523; DRG node-kind #1518; charter/missions paired-invariant gaps #1717/#1718). The trajectory is real and continuous. Qualifier (Directive-003 honesty): the operator's own G1 success criterion is the strict one — "doctrine/DRG demonstrably gates a runtime decision path, with a test proving the directive/contract changes behaviour, not just resolves." The git record shows the plumbing and breadth are there; from the planning POV I can corroborate the trajectory exists and accelerated, but the binary "gates not just resolves" proof is an architect/test-level assertion (naming-identity-ssot-strangler already flags G1 as the next slice, not yet closed). So: continuation confirmed; success-criterion closure pending.

G2 — SUPPORTED (strongest-evidenced)

This is the spine of the release. 28 in-range missions, and the heaviest path-commit concentration in the whole codebase: status 77, core 61, missions 58, coordination 42, lanes 35. The strangler vocabulary is explicit in subjects ("canonical mission_runtime surface + status-facade strangle … epic #1666 slice 2", #1793; "Canonicalize WP lane state machine", #1775; "execution-context unification" #1850). Critically, the last substantive commits before the tag are G2 (#2001 mission-identity naming seam; #1991 coord-topology lanes.json; coordination artifact-home unification) — proving the trajectory runs into 3.2.0, not stopping short of it. The ratchet surface (test_execution_context_parity, test_merge_pipeline_ratchets, test_ratchet_baselines, test_shared_package_boundary) was extended in-range, evidencing the "extract→route→enforce" pattern the goal claims. Net-new pivot? No — these are #1619/#1666/#1868/#1878 epic slices, multi-mission chains.

G3 — SUPPORTED

19 in-range missions plus the largest PR-bucket (37, releng/sonar/ci/lint/hooks/parallel) and 72 workflow-file commits, 760 test-file commits. The two named devex governance artifacts — HOW_TO_MAINTAIN.md and the docs/release-goals/ convention — exist and landed in-range, directly matching the goal text. The seam-extraction-for-testability and ratchet-to-stay-strangled claims are corroborated by the same parity/baseline ratchet files G2 leans on (shared enabler) plus test-suite-acceleration-01KV3H59 (parallelize CI/local, per-worker HOME isolation, #1957). This is a genuine, measurable enabler stream, not a label retrofitted onto incidental cleanup.

Risks / Thin Spots (planner-priti risk register)

  • R1 (classification softness): slug-keyword bucketing has soft edges — merge-* missions could be read as G3-stabilization rather than G2-SSOT; charter-end-user-docs is G1-adjacent but doc-shaped. Counts are directional; the ordinal result (G2 > G1 ≈ G3 ≫ pivot-risk) is robust to reasonable re-bucketing. Mitigation: precedence rule is documented and reproducible.
  • R2 (G1 depth gap): the only place evidence is genuinely thin against the operator's wording is G1's "gates not just resolves" criterion — breadth is proven, the behaviour-gating proof-test is the declared next slice. This is a known-open, not a contradiction.
  • R3 (release noise): ~63 of 119 PRs are release/rc plumbing; if one counts PRs naively, DevEx looks inflated. The mission-level inventory (129 specs) is the cleaner planning signal and is what the verdicts rest on.
  • R4 (corpus vs range): the total kitty-specs corpus (G1 32 / G2 43 / G3 29) shows the same shape as the in-range slice (G1 26 / G2 28 / G3 19), confirming 3.2.x continues a pre-existing trajectory — the goals are continuations, claim VALIDATED.

Bottom Line

The operator's claim — "3.2.0 already moved in these directions; the goals are evidence-grounded continuations, not net-new pivots" — is corroborated from the planning/tracking POV. All three goals map to substantial, dated, multi-mission delivery that runs continuously into the 3.2.0 tag, with G2 as the demonstrable spine. The single honest caveat is G1's strict success-criterion (behaviour-gating proof), which the trajectory approaches but has not yet closed — consistent with the operator's own "cycle stays open" framing.

Edit this page