feat(auto): task-class catalog data (L1-a)#1173
Merged
Merged
Conversation
L1-a slice of Q00#1171 — the ledger-derive redesign of Q00#1157's L1 lane. ## Summary Introduce the 7-class **task-class catalog** as catalog data only. Frozen enum + per-class profile + immutable lookup table. **No inference, no LLM, no eval set, no telemetry** — those belong to later L1-b/c/d sub-PRs and the ledger-derive design explicitly forbids a separate classifier model. The catalog is a *task-class* concept, distinct from the existing ``DomainProfile`` *meta-domain* concept (coding / research / design). Both layers can coexist; this PR introduces the ``task_classes`` module as a sibling, not an extension. ## Why this is small Per the Ouroboros minimal-substrate principle re-emphasized in the 2026-05-22 SSOT self-audit, L1 was scoped down from a classifier + eval set + accuracy floor + opt-in telemetry pipeline to **pattern matching against the interview's standardized ledger**. This PR is the smallest possible substrate that downstream L1-b (the ``derive_domain_from_ledger`` pattern matcher) and L1-c (the Seed AC injection hook) can consume. ## What lands - ``src/ouroboros/auto/task_classes.py`` (new): - ``CompletionMode`` StrEnum: ``CODE_COMPLETE`` / ``PRODUCT_COMPLETE``. - ``TaskClass`` StrEnum: 7 frozen classes (``library``, ``cli``, ``web_service``, ``webhook``, ``data_pipeline``, ``game_2d``, ``refactor_in_place``). Deferred classes (``game_3d``, ``desktop_app``, ``notebook_analysis``) per Q00#1171. - ``TaskClassProfile`` frozen dataclass: ``name``, ``default_completion_mode``, ``default_ac_template`` (``tuple[str, ...]`` matching ``Seed.acceptance_criteria`` exactly), ``runtime_probe_kinds`` (plain-string placeholder until L3 lands). - ``TASK_CLASS_CATALOG``: immutable ``MappingProxyType`` over a private dict so consumers cannot mutate. - ``tests/unit/auto/test_task_classes.py`` (new): - 13 tests covering catalog shape, completion-mode invariants, immutability, StrEnum behavior, and the ``webhook`` vs ``web_service`` distinction pin. ## Test plan - [x] ``uv run pytest tests/unit/auto/test_task_classes.py -v`` → 13 passed. - [x] ``uv run pytest tests/unit/auto -q`` → 889 passed (878 baseline + 13 new − 2 fixture-renames-not-affected = 889). - [x] ``uv run ruff check`` on touched files → clean. - [x] ``uv run ruff format`` on touched files → no changes. - [x] ``uv run mypy src/ouroboros/auto/task_classes.py`` → clean. ## What is NOT in this PR - Pattern-matching ``derive_domain_from_ledger`` — L1-b. - Interview-driver disambiguation hook — L1-c. - Seed Architect AC injection — L1-d (separate file in ``seed_architect``). - Result-envelope surface (``AutoPipelineResult.active_task_class``) — L1-e. ## References - Q00#1157 — Meta SSOT for ``ooo auto`` (L1 lane body, *Substrate honesty* note). - Q00#1171 — L1 design issue (ledger-derive redesign, this PR's spec). - Q00#849 — Original DomainProfile contract (kept untouched; task-class is a parallel concept at a finer granularity). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
5 tasks
![ouroboros-agent[bot]](https://avatars.githubusercontent.com/in/3091302?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Review — ouroboros-agent[bot]
Verdict: REQUEST_CHANGES
Branch: feat/prL1a | 2 files, +325/-0 | CI: Bridge TypeScript pass 13s https://github.com/Q00/ouroboros/actions/runs/26284794948/job/77369602946
Scope: diff-only
HEAD checked: cd49d6e74517bcec330f171f91509c93a2fbfaad
What Improved
- Added a narrow catalog-only L1-a slice with no inference, LLM calls, telemetry, or runtime behavior changes.
- Added focused tests for catalog coverage, profile shape, enum serialization, immutability, completion-mode grouping, and webhook/web-service separation.
Issue #1171 Requirements
| Requirement | Status |
|---|---|
L1-a ships seven classes covering library, cli, web-service, webhook, data-pipeline, game-2d, and refactor-in-place. |
NOT MET — seven enum members exist at src/ouroboros/auto/task_classes.py:72, but several serialized values use underscores rather than the issue’s canonical hyphenated values. |
L1-a includes per-class default_completion_mode, default_ac_template, and runtime_probe_kinds. |
MET — fields are present at src/ouroboros/auto/task_classes.py:107. |
L1-a includes per-class safe_defaults. |
NOT MET — TaskClassProfile ends without this field at src/ouroboros/auto/task_classes.py:109. |
| L1-a includes at least one unit test per class. | MET — tests/unit/auto/test_task_classes.py:33 parametrizes catalog-shape checks across every TaskClass. |
| L1-b/L1-c/L1-d/L1-e inference, interview, AC injection, and result-envelope behavior. | DECLARED OUT OF SCOPE — PR body explicitly defers these later slices; no blocker raised for their absence. |
Prior Findings Status
| Prior Finding | Status |
|---|---|
Previous bot review reported no in-scope blockers but also stated local file reads failed with bwrap: No permissions to create a new namespace. |
MODIFIED — current review inspected the checked-out files and verified current-HEAD contract blockers. |
Blockers
| # | File:Line | Severity | Confidence | Finding |
|---|---|---|---|---|
| 1 | src/ouroboros/auto/task_classes.py:74 |
Medium | 90% | The frozen serialized task-class names use underscores (web_service, with the same pattern at lines 76-78), but the linked L1 design defines the canonical class values as hyphenated names (web-service, data-pipeline, game-2d, refactor-in-place). Because TaskClass is a StrEnum intended to serialize directly into downstream envelopes/ledger-facing values, this locks in a contract that does not match the design issue before L1-b/L1-d consume it. |
| 2 | src/ouroboros/auto/task_classes.py:106 |
Medium | 80% | TaskClassProfile only exposes name, default_completion_mode, default_ac_template, and runtime_probe_kinds through line 109, but the linked L1-a sub-PR breakdown explicitly includes safe_defaults as a per-class field. Since this PR freezes the producer-side catalog schema for later inference and Seed assembly, omitting that field leaves the L1-a contract incomplete rather than merely deferred. |
Follow-ups
| # | File:Line | Priority | Confidence | Suggestion |
|---|
Test Coverage
tests/unit/auto/test_task_classes.py:26 and tests/unit/auto/test_task_classes.py:33 cover enum/catalog lockstep and profile shape for every class. However, tests currently pin the underscore serialized values at tests/unit/auto/test_task_classes.py:93, so they do not catch the canonical-name mismatch. They also cannot cover safe_defaults because the field is absent from TaskClassProfile.
I attempted uv run pytest tests/unit/auto/test_task_classes.py -q, but the run failed before tests executed because the editable build’s VCS version hook timed out on git version and then could not derive a version. A direct PYTHONPATH=src import smoke check for the new module passed.
Design / Roadmap Gate
The PR aligns with the design gate on keeping L1-a catalog-only and avoiding a separate classifier, eval set, telemetry, or new LLM path. It does not fully align with the L1-a catalog contract because current HEAD freezes underscore serialized class values and omits the safe_defaults field that the linked issue lists for the per-class schema.
Merge Recommendation
- Merge after the catalog’s serialized names and profile schema are brought back into alignment with the L1-a design contract, with tests updated to pin those contract values.
ouroboros-agent[bot]
Two docstring-only clarifications surfaced by code review on PR Q00#1173: 1. Explain why `safe_defaults` is intentionally absent from `TaskClassProfile`. The earlier Q00#1171 schema sketch carried that field forward from Q00#849's `DomainProfile`, but on implementation review it is meta-domain-scoped (applies uniformly across coding task classes) and belongs on the `DomainProfile` layer, not on the within-meta-domain task-class catalog. The decoupling rationale in the module docstring now lists the full set of meta-domain fields kept on `DomainProfile` and notes the deliberate omission. 2. Note that catalog serialization uses underscored identifiers (`web_service`, `data_pipeline`, `game_2d`, `refactor_in_place`) so `TaskClass.value` is a valid Python identifier and JSON-safe ledger key. Prose docs in the Q00#1157 / Q00#1171 issues may render the same names with hyphens; both refer to the same class. This prevents L1-b pattern functions from hardcoding the wrong form. No behavior change. Tests, lint, mypy unchanged.
Collaborator
Author
Merge JustificationThis PR is the L1-a slice of #1171 (DomainProfile catalog), which is in turn a slice of #1157 (the What this PR isA catalog-data-only module (
No inference logic, no LLM, no eval set, no telemetry, no caller wiring — those land in L1-b/c/d/e per the issue's sub-PR breakdown. Why this aligns with the SSOTs
What changed in this iterationOne follow-up commit on top of
Why over-engineering risk is low
Verification
This PR is small, targeted, well-tested, aligned with the SSOT-confirmed direction, and unblocks L1-b/c/d/e cleanly. Recommending merge. |
Contributor
There was a problem hiding this comment.
Review — ouroboros-agent[bot]
Verdict: APPROVE
Reviewing commit
9aae198for PR #1173
Review record:
854cc3f0-6b25-4e02-b2ac-8f4e6ecaa7c8
Blocking Findings
No in-scope blocking findings remained after policy filtering.
Non-blocking Suggestions
None.
Design Notes
Review could not be completed due to the execution environment blocking all local file reads.
Policy Notes
- Omitted 1 finding(s) that referenced files outside the current PR changed-files scope.
Recovery Notes
First recoverable review artifact generated from codex analysis log.
Reviewed by ouroboros-agent[bot] via Codex deep analysis
Collaborator
Author
PR Review SummaryVerdictAPPROVE Scope Reviewed
Blocking IssuesNone. WarningsNone. Mutation-Test Thinking
Complexity / CRAP-style Risk
Test Quality Assessment
Security / Operational RiskNone. The module is import-time data initialization only: no I/O, no network, no subprocess, no logging, no secrets, no auth, no user-input parsing, no migrations, no callers. Adding the module cannot regress any production code path because nothing in SSOT Alignment Check
Looks Good
Final RecommendationAPPROVE — ready to merge. The PR is small (catalog data only), aligned with both governing SSOTs (#1157 and #1171), tested at the mutation-survivor level for every structural invariant it promises, and carries zero security or operational risk because it has no callers and no I/O. The single follow-up commit |
Contributor
There was a problem hiding this comment.
Review — ouroboros-agent[bot]
Verdict: APPROVE
Branch: feat/prL1a | 2 files, +339/-0 | CI: Bridge TypeScript pass 14s https://github.com/Q00/ouroboros/actions/runs/26288809936/job/77383216437
Scope: diff-only
HEAD checked: 9aae19832b7c2c8020afd1866e13ac48453ca5c5
What Improved
- Added a narrow catalog-only L1-a module with no runtime wiring, classifier, telemetry, or LLM path.
- Added focused unit tests for enum/catalog lockstep, profile shape, completion-mode partitioning, immutability, StrEnum serialization, and webhook/web-service separation.
Issue #1171 Requirements
| Requirement | Status |
|---|---|
| L1-a ships seven classes: library, CLI, web service, webhook, data pipeline, game 2D, refactor in place. | MET — enum values are present at src/ouroboros/auto/task_classes.py:77. |
| Catalog provides per-class completion mode, default AC template, and runtime probe kind hints. | MET — TaskClassProfile defines these fields at src/ouroboros/auto/task_classes.py:120, with catalog entries at src/ouroboros/auto/task_classes.py:141. |
| L1-a includes at least one unit test per class. | MET — tests/unit/auto/test_task_classes.py:33 parametrizes checks over list(TaskClass). |
| No classifier, no LLM, no eval set, no telemetry in L1-a. | MET — changed code is catalog data only; no runtime caller imports it, and the module states this boundary at src/ouroboros/auto/task_classes.py:10. |
| L1-b/L1-c/L1-d/L1-e inference, interview, AC injection, and result-envelope behavior. | DECLARED OUT OF SCOPE — PR body defers these later slices; no current runtime behavior is changed. |
Per-class safe_defaults from the issue schema sketch. |
DECLARED NON-GOAL / PRESERVED — current HEAD documents the task-class/domain-profile split at src/ouroboros/auto/task_classes.py:19, while existing DomainProfile.safe_defaults remains intact at src/ouroboros/auto/domain_profile.py:148. |
Prior Findings Status
| Prior Finding | Status |
|---|---|
prev_review.txt reported no in-scope blockers but also said implementation artifacts were inaccessible. |
MODIFIED — current review inspected current HEAD files directly and verified the catalog/tests. |
| Prior artifact blocker: underscored serialized task-class names diverge from hyphenated issue prose. | WITHDRAWN — current HEAD explicitly documents the serialization convention at src/ouroboros/auto/task_classes.py:38; no current consumer exists, and tests pin the intended StrEnum values at tests/unit/auto/test_task_classes.py:90. |
Prior artifact blocker: TaskClassProfile omits per-class safe_defaults. |
WITHDRAWN — current HEAD declares safe_defaults intentionally remains on the existing meta-domain DomainProfile layer at src/ouroboros/auto/task_classes.py:19, and the existing field remains present at src/ouroboros/auto/domain_profile.py:148. |
Blockers
| # | File:Line | Severity | Confidence | Finding |
|---|
Follow-ups
| # | File:Line | Priority | Confidence | Suggestion |
|---|---|---|---|---|
| 1 | src/ouroboros/auto/task_classes.py:10 |
Low | 80% | The module docstring’s follow-up labels are one step off from issue #1171’s sub-PR breakdown: it names Seed AC injection as L1-c and result-envelope as L1-d, while #1171 lists interview integration as L1-c, Seed AC injection as L1-d, and result-envelope as L1-e. This is traceability-only and does not affect the catalog contract. |
Test Coverage
tests/unit/auto/test_task_classes.py:26 covers enum/catalog lockstep, tests/unit/auto/test_task_classes.py:33 parametrizes shape checks across every TaskClass, tests/unit/auto/test_task_classes.py:58 pins the CODE_COMPLETE partition, and tests/unit/auto/test_task_classes.py:98 pins the webhook/web-service split. All newly added catalog logic and state-shape invariants have corresponding tests for this data-only slice.
Verification run: SETUPTOOLS_SCM_PRETEND_VERSION=0.0.0 uv run pytest tests/unit/auto/test_task_classes.py -q passed, ruff check passed, and mypy src/ouroboros/auto/task_classes.py passed. Plain uv run pytest ... failed before tests because this environment’s default git wrapper timed out during VCS version detection, not because of test failures.
Design / Roadmap Gate
The PR aligns with the L1-a catalog-only gate in design_context.md: it preserves the no-classifier/no-telemetry ledger-derived direction and defers inference, interview integration, Seed AC injection, and result-envelope plumbing. The only design wrinkle is traceability wording around follow-up slice numbering in the new docstring; it is non-blocking because the changed implementation remains data-only and does not alter runtime contracts.
Merge Recommendation
- Ready to merge; the one follow-up is documentation traceability cleanup and does not need to block.
ouroboros-agent[bot]
shaun0927
added a commit
to shaun0927/ouroboros
that referenced
this pull request
May 22, 2026
Resolve the two README/code mismatches surfaced by the latest review on PR Q00#1174: 1. `tests/canonical/README.md` "Full live run" section claimed `OUROBOROS_RUN_CANONICAL=1 uv run pytest tests/canonical/ -v` actually invokes `ouroboros_auto` against each scenario. In L0-a the live wiring is deferred — `test_scenario_live_run_or_skip` unconditionally `pytest.skip`s with a typed reason after the opt- in env var is checked, so the documented invocation behavior is not yet available. Reword the section to call out the L0-a state ("opt-in still skips with a typed reason; shape-check tests still run") while keeping the future L0-b semantics described. 2. `tests/canonical/README.md` "Run a single scenario" section pointed at `tests/canonical/cli-todo/`, but the scenario directory contains only `goal.txt` and `expected.yaml` — pytest collects zero tests there. The actual test bodies live in `tests/canonical/test_canonical.py` and are parametrized per scenario via `pytest_generate_tests`. Replace the command with the working filter form: `uv run pytest tests/canonical/ -v -k <slug>`. 3. `tests/canonical/cli-todo/expected.yaml` had a comment referencing `test_scenario_completion_mode_matches_catalog`, which is not in the harness and is deferred until Q00#1173 (L1-a catalog data) is available on `main`. Update the comment to note the round-trip test is a follow-up, not yet present. No code change. The hermetic shape-check suite is unchanged (still 6 passed, 1 skipped). `uv run pytest tests/canonical/ -v -k cli-todo` now collects and passes the per-scenario tests, replacing the previously documented zero-collection command.
This was referenced May 22, 2026
shaun0927
added a commit
to shaun0927/ouroboros
that referenced
this pull request
May 22, 2026
Align the L1-a catalog documentation with Q00#1171's current sub-PR breakdown so future readers do not treat Seed AC injection as L1-c or result-envelope plumbing as L1-d. Constraint: PR Q00#1173 is already merged; follow-up must be docs-only and preserve the approved data-only catalog behavior. Rejected: Change catalog schema or serialized task-class values | ouroboros-agent already approved those after the decoupling rationale, and the remaining bot note was traceability-only. Confidence: high Scope-risk: narrow Directive: Keep L1-c reserved for interview-driver disambiguation, L1-d for Seed AC injection, and L1-e for result-envelope plumbing unless Q00#1171 is explicitly revised. Tested: SETUPTOOLS_SCM_PRETEND_VERSION=0.0.0 uv run pytest tests/unit/auto/test_task_classes.py -q; uv run ruff check src/ouroboros/auto/task_classes.py tests/unit/auto/test_task_classes.py; uv run ruff format --check src/ouroboros/auto/task_classes.py tests/unit/auto/test_task_classes.py; SETUPTOOLS_SCM_PRETEND_VERSION=0.0.0 uv run mypy src/ouroboros/auto/task_classes.py Not-tested: Full unit suite; change is limited to comments/docstrings. Co-authored-by: OmX <omx@oh-my-codex.dev>
This was referenced May 22, 2026
This was referenced May 23, 2026
Contributor
There was a problem hiding this comment.
Review — ouroboros-agent[bot]
Verdict: APPROVE
PR #1173
Branch: feat/prL1a | 2 files, +339/-0 | CI: Bridge TypeScript pass 14s https://github.com/Q00/ouroboros/actions/runs/26288809936/job/77383216437
Scope: architecture-level
HEAD checked: 9aae19832b7c2c8020afd1866e13ac48453ca5c5
What Improved
- Added a focused
ouroboros.auto.task_classescatalog with explicit task-class and completion-mode enums. - Kept the L1-a surface data-only: no inference, no persistence migration, no runtime probe execution, and no Seed mutation hook landed in this slice.
- Added unit coverage for catalog shape, enum/catalog parity, direct immutability of the public mapping view, StrEnum serialization behavior, and the webhook/web-service distinction.
Issue #N/A Requirements
| Requirement | Status |
|---|---|
| Add 7-class task-class catalog | Satisfied |
Add CompletionMode StrEnum with code/product completion modes |
Satisfied |
Add frozen TaskClassProfile with name, default completion mode, default AC template, and runtime probe hints |
Satisfied |
| Keep catalog data-only with no inference, LLM, eval set, telemetry, Seed injection, or result-envelope changes | Satisfied |
Keep task classes separate from DomainProfile meta-domain layer |
Satisfied |
| Use immutable public catalog view | Satisfied for public mapping mutation |
| Add meaningful tests for newly added logic | Satisfied for this data-only slice |
Prior Findings Status
| Prior Finding | Status |
|---|---|
| Prior review context | MODIFIED — Prior concerns were modified/withdrawn for current HEAD: the current file documents the safe-defaults decoupling rationale and the underscore serialization convention, and I found no remaining current-HEAD blocker with file:line evidence. |
Blockers
| # | File:Line | Severity | Confidence | Finding |
|---|
Follow-ups
| # | File:Line | Priority | Confidence | Suggestion |
|---|---|---|---|---|
| — | — | — | — | None. |
Test Coverage
- Verified
uv run pytest tests/unit/auto/test_task_classes.py -q: 13 passed. - Verified
uv run ruff check src/ouroboros/auto/task_classes.py tests/unit/auto/test_task_classes.py: passed. - Verified
uv run mypy src/ouroboros/auto/task_classes.py: passed. - Coverage is appropriately scoped for this data-only L1-a slice; integration tests for ledger derivation, Seed AC injection, result-envelope persistence, replay, and runtime probes remain correctly deferred because those behaviors are not implemented in current HEAD.
Design / Roadmap Gate
Affected-boundary review covered the new catalog API, Seed.acceptance_criteria shape, DomainProfile separation, auto package export patterns, and downstream persistence/replay surfaces. Because current HEAD only introduces static catalog data and does not wire task class into state, ledger derivation, Seed mutation, result envelopes, or runtime probes, there is no new state-machine, persistence, replay, or consumer-contract blocker visible at current HEAD.
Merge Recommendation
Post-merge audit only: no current-HEAD blocker found for the landed L1-a catalog slice. Keep the follow-on L1-b/L1-c/L1-d work gated on tests that exercise derivation, persistence/replay compatibility, Seed AC injection, and result-envelope consumer behavior.
Review-Metadata:
verdict: APPROVE
github_event: COMMENT
review_kind: post_merge_audit
merge_eligible: false
head_sha: 9aae198
source_read_ok: true
diff_read_ok: true
blocking_count: 0
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
L1-a slice of #1171 — the ledger-derive redesign of #1157's L1 lane.
Introduces the 7-class task-class catalog as catalog data only. Frozen enum + per-class profile + immutable lookup table. No inference, no LLM, no eval set, no telemetry — those belong to later L1-b/c/d sub-PRs and the ledger-derive design explicitly forbids a separate classifier model.
The catalog is a task-class concept, distinct from the existing
DomainProfilemeta-domain concept (coding / research / design). Both layers can coexist; this PR introduces thetask_classesmodule as a sibling, not an extension.Why this is small
Per the Ouroboros minimal-substrate principle re-emphasized in the 2026-05-22 SSOT self-audit (#1157), L1 was scoped down from a classifier + eval set + accuracy floor + opt-in telemetry pipeline to pattern matching against the interview's standardized ledger. This PR is the smallest possible substrate that downstream L1-b (the
derive_domain_from_ledgerpattern matcher) and L1-c (the Seed AC injection hook) can consume.What lands
src/ouroboros/auto/task_classes.py(new):CompletionModeStrEnum:CODE_COMPLETE/PRODUCT_COMPLETE.TaskClassStrEnum: 7 frozen classes (library,cli,web_service,webhook,data_pipeline,game_2d,refactor_in_place). Deferred classes (game_3d,desktop_app,notebook_analysis) per Meta SSOT slice: L1 — TaskClass Catalog (ledger-derived domain inference + default AC injection) #1171.TaskClassProfilefrozen dataclass:name,default_completion_mode,default_ac_template(tuple[str, ...]matchingSeed.acceptance_criteriaexactly),runtime_probe_kinds(plain-string placeholder until L3 lands).TASK_CLASS_CATALOG: immutableMappingProxyTypeover a private dict.tests/unit/auto/test_task_classes.py(new): 13 tests covering catalog shape, completion-mode invariants, immutability, StrEnum behavior, webhook-vs-web-service distinction pin.What is NOT in this PR
derive_domain_from_ledger— L1-b.AutoPipelineResult.active_task_class) — L1-e.Test plan
uv run pytest tests/unit/auto/test_task_classes.py -v→ 13 passed.uv run pytest tests/unit/auto -q→ 889 passed.uv run ruff checkon touched files → clean.uv run ruff formaton touched files → no changes.uv run mypy src/ouroboros/auto/task_classes.py→ clean.Refs #1157 (L1 lane), #1171 (L1 design issue), #849 (existing DomainProfile contract — preserved untouched).