After Step 13

This commit is contained in:
GAJ3T 2026-11-01 10:19:57 -05:00
parent 18c9d0d65b
commit 13221a94a9
13 changed files with 1115 additions and 71 deletions

View File

@ -170,6 +170,10 @@ ownership are separate from proposal and programming evidence.
See [comparison usage](docs/calibration-comparison.md) and See [comparison usage](docs/calibration-comparison.md) and
[Step-13 worker evidence](prompts/Step-13-Report.md). Step-13 requires manual [Step-13 worker evidence](prompts/Step-13-Report.md). Step-13 requires manual
acceptance; Step-14 remains blocked. acceptance; Step-14 remains blocked.
Step-13A corrects the remaining owner-release tail: Busy and the exact admitted
drain stay reserved through cleanup, before atomic terminal completion.
See [corrective execution evidence](prompts/Step-13A-Report.md). Step-13/13A
remain pending manual acceptance; no model or programming scope is added.
### Planned ### Planned

View File

@ -28,7 +28,8 @@ pending/manual/stale/units/source checks. All 596 tests pass (594 retained plus
two corrective cases). Step-12/12A is accepted. Step-13 adds an explicit two corrective cases). Step-12/12A is accepted. Step-13 adds an explicit
verified-image-to-NEW-configuration equivalent synthetic comparison, bounded verified-image-to-NEW-configuration equivalent synthetic comparison, bounded
real captures, canonical metrics/lineage and existing save/readback/replay. real captures, canonical metrics/lineage and existing save/readback/replay.
Step-13 awaits manual review; Step-14 and physical execution remain blocked. Step-13A corrects the comparison release/completion tail while preserving that
scope. Step-13/13A await manual review; Step-14 and physical execution remain blocked.
- [Architecture](architecture.md): actual nine-project graph, responsibilities, - [Architecture](architecture.md): actual nine-project graph, responsibilities,
route ownership and provider/reference boundaries. route ownership and provider/reference boundaries.

View File

@ -304,7 +304,12 @@ One retained result plus bounded partial run/save evidence is inspectable.
New evaluation or Close retires earlier current status; source/head/target New evaluation or Close retires earlier current status; source/head/target
supersession also computes Stale without rebase. Hybrid events release before supersession also computes Stale without rebase. Hybrid events release before
admitted blocking work. Cancel/Close drains outside required gates; resources admitted blocking work. Cancel/Close drains outside required gates; resources
dispose even if capture cleanup fails. Explicit saves use existing no-overwrite dispose even if capture cleanup fails. Step-13A keeps the actual owner/busy/drain
reserved throughout release, then settles terminal state, owner removal and
that exact completion atomically under the comparison gate. Callback/release
work never runs under that state gate; cleanup errors preserve the primary
operation error and settle waiters as failures rather than exposing premature
idle. Shared Close cannot miss an earlier unfinished owner. Explicit saves use existing no-overwrite
archive storage, exact serialized readback and per-ID outcomes. A failed save archive storage, exact serialized readback and per-ID outcomes. A failed save
may have published: retry checks the existing catalogue and verifies read-only may have published: retry checks the existing catalogue and verifies read-only
instead of colliding or overwriting. No compensating file deletion occurs. instead of colliding or overwriting. No compensating file deletion occurs.

View File

@ -48,6 +48,7 @@ dotnet test .\src\tests\AutoForge.IntegrationTests\AutoForge.IntegrationTests.cs
dotnet test .\src\tests\AutoForge.IntegrationTests\AutoForge.IntegrationTests.csproj --configuration Release --no-restore --filter "FullyQualifiedName~RunbookPreviewTests|FullyQualifiedName~RunbookUi_|FullyQualifiedName~RunbookIsolation_" dotnet test .\src\tests\AutoForge.IntegrationTests\AutoForge.IntegrationTests.csproj --configuration Release --no-restore --filter "FullyQualifiedName~RunbookPreviewTests|FullyQualifiedName~RunbookUi_|FullyQualifiedName~RunbookIsolation_"
dotnet test .\src\tests\AutoForge.IntegrationTests\AutoForge.IntegrationTests.csproj --configuration Release --no-restore --filter "FullyQualifiedName~LocalRunbookGenerationTests|FullyQualifiedName~LocalInferenceBoundsTests|FullyQualifiedName~LocalGenerationReviewTests|FullyQualifiedName~LocalGenerationUi_|FullyQualifiedName~LocalGenerationIsolation_|FullyQualifiedName~RunbookPreviewTests" dotnet test .\src\tests\AutoForge.IntegrationTests\AutoForge.IntegrationTests.csproj --configuration Release --no-restore --filter "FullyQualifiedName~LocalRunbookGenerationTests|FullyQualifiedName~LocalInferenceBoundsTests|FullyQualifiedName~LocalGenerationReviewTests|FullyQualifiedName~LocalGenerationUi_|FullyQualifiedName~LocalGenerationIsolation_|FullyQualifiedName~RunbookPreviewTests"
dotnet test .\src\tests\AutoForge.IntegrationTests\AutoForge.IntegrationTests.csproj --configuration Release --no-restore --filter "FullyQualifiedName~Comparison_|FullyQualifiedName~PreviewHandoff_|FullyQualifiedName~TorqueDraft_" dotnet test .\src\tests\AutoForge.IntegrationTests\AutoForge.IntegrationTests.csproj --configuration Release --no-restore --filter "FullyQualifiedName~Comparison_|FullyQualifiedName~PreviewHandoff_|FullyQualifiedName~TorqueDraft_"
dotnet test .\src\tests\AutoForge.IntegrationTests\AutoForge.IntegrationTests.csproj --configuration Release --no-restore --filter "FullyQualifiedName~ComparisonOwnershipTail_|FullyQualifiedName~ComparisonOwnerCleanup_|FullyQualifiedName~Comparison_|FullyQualifiedName~PreviewHandoff_|FullyQualifiedName~TorqueDraft_"
``` ```
## Step-12A real local-model/native qualification ## Step-12A real local-model/native qualification
@ -106,6 +107,30 @@ exposed the unchanged recording panel's pending-refresh Busy initialization
error. Task-owned Client recovery and sequential exact-heading/idle-archive-owner error. Task-owned Client recovery and sequential exact-heading/idle-archive-owner
checks passed for all six existing routes. No recording-navigation change was made. checks passed for all six existing routes. No recording-navigation change was made.
## Step-13A ownership-tail correction
The actual pre-fix selector `FullyQualifiedName~ComparisonOwnershipTail_`
executed seven cases: all seven failed behaviorally, zero passed/skipped.
Real cancellation callbacks held the old release after actual save work
returned: Busy was false, competing Evaluate/Save admitted and Close/Dispose
could complete with the old drain/callback unfinished. A cancelled actual
evaluation also exposed premature terminal publication.
The correction releases outside the state gate while retaining owner/busy/drain,
then atomically publishes terminal state/owner removal/completion. Nine new
cases cover successful/failed/cancelled work, actual delayed callback cleanup,
competing admission, orphan prevention, exact drain identity, legitimate retry/
fresh lifetime and cleanup failure aggregation/per-ID save preservation.
The combined selector above passes 59 cases; final sequential nine-project
Release build passes with zero warnings/errors, followed by 179 unit + 475
integration = 654 passes, zero failed/skipped. All 645 Step-13 cases remain.
Affected native lifecycle and one legitimate actual comparison/save/report are
verified separately from controlled tail barriers in the
[Step-13A report](../prompts/Step-13A-Report.md). Prior real LLM content and the
original 645-case/numerical/archive evidence remain historical, not rerun claims.
Step-13/13A await manual acceptance. No unrelated recording-navigation/dynamics
fix, runtime setup, new model/physics/schema or Step-14 work is included.
## Historical Step-12 approval-blocked verification ## Historical Step-12 approval-blocked verification
Step-11 is accepted. Step-12 implementation/control tests pass, but required Step-11 is accepted. Step-12 implementation/control tests pass, but required

View File

@ -1,6 +1,6 @@
# Verified synthetic calibration comparison # Verified synthetic calibration comparison
Step-13 is implemented pending manual review. This is a narrow synthetic Step-13 with corrective Step-13A is implemented pending manual review. This is a narrow synthetic
comparison, not manufacturer prediction, safe tuning advice, firmware proof, comparison, not manufacturer prediction, safe tuning advice, firmware proof,
physical authority or a general AI/runbook executor. physical authority or a general AI/runbook executor.
@ -94,6 +94,20 @@ Close retires earlier current status; source/head/proof supersession labels it
Stale/as-of without rebase. One previous success and bounded latest partial run/ Stale/as-of without rebase. One previous success and bounded latest partial run/
save evidence remain inspectable, not an unbounded history. save evidence remain inspectable, not an unbounded history.
Busy includes the entire completion tail, not merely returned operation work.
The admitted owner and its exact drain remain reserved while cancellation
callbacks/release finish outside the comparison state gate. Only then are
terminal state, owner removal and that completion settled together under the
gate. A competing Evaluate/Save cannot replace the old drain; shared Close/
Dispose includes all admitted cleanup. After a full drain, legitimate new work
can admit only while open; closed sessions require an independent new lifetime.
Cleanup errors fault operation waiters explicitly, preserving primary error/
cancellation and per-capture save evidence; combined failures retain both errors.
Controlled tests use a minimal internal real-release observation seam, not a
public test control or mocked coordinator. Native UI uses the same unmodified
commands with no lifecycle probe subscribed. See the
[Step-13A report](../prompts/Step-13A-Report.md) for pre-fix and corrected evidence.
Failed/limited/cancelled physics never becomes a successful pair. A stopped Failed/limited/cancelled physics never becomes a successful pair. A stopped
capture can represent complete coverage of a partial run; the run's separate capture can represent complete coverage of a partial run; the run's separate
error/metrics/completion determine comparison success. This does not expand or error/metrics/completion determine comparison success. This does not expand or

View File

@ -3,11 +3,11 @@
## Status and authority ## Status and authority
- Initialized: 2026-10-28. - Initialized: 2026-10-28.
- Current phase: Step-13 comparison reviewed; owner/drain tail requires focused correction. - Current phase: corrected synthetic demonstration accepted; bounded hardware readiness research next.
- Active assignment: [Step-13A](Step-13A.md), ready for manual worker handoff. - Active assignment: [Step-14](Step-14.md), ready for manual worker handoff; research checkpoint only.
- Last accepted milestone: Step-12 with Step-12A, accepted 2026-10-31. - Last accepted milestone: Step-13 with Step-13A, accepted 2026-11-01.
- Last accepted correction/follow-up: Step-12A; all frozen gates closed. - Last accepted correction/follow-up: Step-13A; R-01 closed.
- Outstanding follow-up: Step-13A / R-01 original C-07 owner/drain completion; no Step-12B. - Outstanding follow-up: none; no Step-13B required.
- Step-05 status: ACCEPTED after Step-05A/Step-05B, 2026-10-29. - Step-05 status: ACCEPTED after Step-05A/Step-05B, 2026-10-29.
- Step-06 status: ACCEPTED after Step-06A, 2026-10-30. - Step-06 status: ACCEPTED after Step-06A, 2026-10-30.
- Step-07 status: ACCEPTED after Step-07A/Step-07B, 2026-10-30. - Step-07 status: ACCEPTED after Step-07A/Step-07B, 2026-10-30.
@ -19,9 +19,9 @@
- Step-12A status: ACCEPTED 2026-10-31; worker reports 596 passes, clean Release and actual native exact-intent/lifecycle proof. - Step-12A status: ACCEPTED 2026-10-31; worker reports 596 passes, clean Release and actual native exact-intent/lifecycle proof.
- Step-12 completion prompt: superseded for execution by owner-requested Step-12A; preserve unchanged. - Step-12 completion prompt: superseded for execution by owner-requested Step-12A; preserve unchanged.
- Owner runtime policy: local models only; no Ollama cloud or cloud fallback. - Owner runtime policy: local models only; no Ollama cloud or cloud fallback.
- Step-13 status: NEEDS CORRECTION / C-07 and AC-03; native/645-case evidence reviewed. - Step-13 status: ACCEPTED 2026-11-01 after Step-13A; complete supported synthetic demonstration.
- Step-13A status: READY FOR MANUAL WORKER HANDOFF; one ownership-tail correction only. - Step-13A status: ACCEPTED 2026-11-01; worker reports 654 passes, clean Release and rebuilt affected native proof.
- Step-14 status: read-only hardware; BLOCKED / NOT ASSIGNED until demonstration acceptance and hardware gates. - Step-14 status: READY FOR MANUAL WORKER HANDOFF / READINESS ONLY; actual physical integration unimplemented and separately gated.
- Current state and handoff: [STATE.md](STATE.md). - Current state and handoff: [STATE.md](STATE.md).
- Authority: the master charter, repository/handoff amendment, and the owner's - Authority: the master charter, repository/handoff amendment, and the owner's
2026-10-28 architectural realignment. The latest decision cancels the dual-shell 2026-10-28 architectural realignment. The latest decision cancels the dual-shell
@ -84,7 +84,7 @@ and `Step-01A-Report.md`, then `Step-01B.md`. Accept all necessary corrections
before assigning the next major step. Do not use the superseded numeric-dot before assigning the next major step. Do not use the superseded numeric-dot
prompt/report naming convention. prompt/report naming convention.
## Assignment sizing and handoff policy - revised 2026-10-28 ## Assignment sizing and handoff policy - revisions 2026-10-28 and 2026-11-01
The owner requested an orchestration-level reassessment to reduce repeated The owner requested an orchestration-level reassessment to reduce repeated
corrective handoffs. Longer prose and larger passing test counts have not been corrective handoffs. Longer prose and larger passing test counts have not been
@ -127,6 +127,28 @@ retroactively to completed assignments.
can guarantee zero defects; the goal is fewer preventable handoffs, not weaker can guarantee zero defects; the goal is fewer preventable handoffs, not weaker
review or concealed failures. review or concealed failures.
### Stronger prospective control from Step-14
The owner repeated the request after Step-13A. Its R-01 escaped because the
pending-work gate/self-review did not explicitly hold the post-work release/
drain tail. The orchestrator also needs smaller prerequisite-first assignments,
not a broad new adapter/protocol/UI/driver state machine in one handoff.
- Revised shared gate requires source-ordering/evidence ledger for admission,
publication, completion/close and affected real UI paths. Cleanup-tail tests
hold work after it returns, with exact owner/drain identity and no new calls.
- Worker reuses accepted prior art, fixes tightly coupled findings before
COMPLETE and validates the final changed production/native path. Counts and
prose cannot replace a missing executed criterion.
- Apply only relevant rows and record N/A concretely; no extra checklist files,
agents, retrospective new criteria or blanket auditing of unaffected code.
- Orchestrator freezes decisive criteria before manual handoff and reviews
those same rows in one consolidated pass. Separate unverified prerequisites
from implementation and from actual corrective defects.
- Step-14 starts with research/readiness, not source scaffolding or hardware
traffic. Review exact supported initial adapter contract before assigning
code; research completion is not full hardware acceptance or Step-15 entry.
Unassigned future rows remain milestone envelopes, not a requirement to implement Unassigned future rows remain milestone envelopes, not a requirement to implement
their entire scope in one worker turn. If an envelope is too large, split and their entire scope in one worker turn. If an envelope is too large, split and
update the future major-step numbering/dependencies before handoff; letters update the future major-step numbering/dependencies before handoff; letters
@ -596,9 +618,9 @@ evidence; they must not silently discard required capabilities.
| 11 | Deterministic simulation runbook validation and preview | 10 and corrections accepted | ACCEPTED 2026-10-31 | Strict schema/limits/allowlist/source-head/template, immutable preview/diff/staleness and no execution; reported 179 unit/368 integration passes | | 11 | Deterministic simulation runbook validation and preview | 10 and corrections accepted | ACCEPTED 2026-10-31 | Strict schema/limits/allowlist/source-head/template, immutable preview/diff/staleness and no execution; reported 179 unit/368 integration passes |
| 12 | Local AI and applicable knowledge integration | 11 accepted | ACCEPTED 2026-10-31 AFTER 12A | Actual local-only original model/native preview and exact intent/lifecycle; immutable validator/authority boundaries, reported 596 passes | | 12 | Local AI and applicable knowledge integration | 11 accepted | ACCEPTED 2026-10-31 AFTER 12A | Actual local-only original model/native preview and exact intent/lifecycle; immutable validator/authority boundaries, reported 596 passes |
| 12A | Local-only Ollama setup and real-model qualification | 12 returned incomplete; owner directive | ACCEPTED 2026-10-31 | Cloud-disabled true/current model proof, exact native +600/-10 output, pending conflicts/Cancel/unmount and reproduced responsive Hybrid dispatch fixes | | 12A | Local-only Ollama setup and real-model qualification | 12 returned incomplete; owner directive | ACCEPTED 2026-10-31 | Cloud-disabled true/current model proof, exact native +600/-10 output, pending conflicts/Cancel/unmount and reproduced responsive Hybrid dispatch fixes |
| 13 | Comparative evaluation and complete synthetic demonstration | 12 and 12A accepted | NEEDS CORRECTION AFTER REVIEW | Most proof/projection/native/run/archive/645-case evidence supported; C-07 owner/drain tail replacement remains | | 13 | Comparative evaluation and complete synthetic demonstration | 12 and 12A accepted | ACCEPTED 2026-11-01 AFTER 13A | Trusted projection/equivalent real runs/captures/metrics/native demonstration and corrected completion-tail ownership; reported 654 passes |
| 13A | Retain comparison ownership through cleanup and drain completion | 13 review R-01 / C-07 | READY FOR MANUAL WORKER HANDOFF | Real pre-fix tail-cleanup/reentry/Close proof, owner release outside gates followed by atomic terminal owner/drain publication; supported comparison behavior preserved | | 13A | Retain comparison ownership through cleanup and drain completion | 13 review R-01 / C-07 | ACCEPTED 2026-11-01 | Seven actual pre-fix failures/nine corrected cases; callback/release outside gate, reserved owner then atomic terminal/drain/Close, rebuilt native evidence |
| 14 | Verified OBDX Pro GT read-only integration | 13 plus documented device research/access | BLOCKED / NOT ASSIGNED; HARDWARE GATE | Source-backed device support matrix, explicit physical target, real read-only identification/diagnostics/telemetry recording where supported; no write path | | 14 | Verified OBDX Pro GT read-only integration | 13 and 13A accepted; research before implementation | READY FOR READINESS CHECKPOINT; INTEGRATION GATED | Exact attributable GT protocol/interface/Windows/API/licensing/device-readiness matrix and smallest read-only contract first; no hardware/source implementation in current assignment |
| 15 | Advanced dashboards, analysis, and comparisons | 13; 14 for real captures | PLANNED | Persist configurable layouts/thresholds, linked charts, larger-capture analysis, raw-data inspection and multi-run/multi-vehicle comparisons with lineage/uncertainty | | 15 | Advanced dashboards, analysis, and comparisons | 13; 14 for real captures | PLANNED | Persist configurable layouts/thresholds, linked charts, larger-capture analysis, raw-data inspection and multi-run/multi-vehicle comparisons with lineage/uncertainty |
| 16 | Broader diagnostic protocols and hardware adapters | 14 | PLANNED | Conformance-backed additional CAN/ISO-TP/UDS/KWP2000/J2534 or legacy adapters where verified; same application ports and capability restrictions | | 16 | Broader diagnostic protocols and hardware adapters | 14 | PLANNED | Conformance-backed additional CAN/ISO-TP/UDS/KWP2000/J2534 or legacy adapters where verified; same application ports and capability restrictions |
| 17 | Curated vehicle knowledge and research workflow | 12; 16 where needed | PLANNED | Sourced profile/module/calibration knowledge, review/promotion workflow, exact applicability, licensing/access controls; research cannot authorize writes | | 17 | Curated vehicle knowledge and research workflow | 12; 16 where needed | PLANNED | Sourced profile/module/calibration knowledge, review/promotion workflow, exact applicability, licensing/access controls; research cannot authorize writes |
@ -694,19 +716,20 @@ evidence; they must not silently discard required capabilities.
validator; actual native intent/owner and unavailable/invalid/isolation gates validator; actual native intent/owner and unavailable/invalid/isolation gates
verified. Broader diagnostic/scenario planning, retrieval and restricted tools verified. Broader diagnostic/scenario planning, retrieval and restricted tools
remain planned, not implemented or authority to issue hardware frames. remain planned, not implemented or authority to issue hardware frames.
- **13 assigned:** trusted verified supported-image admission, explicit new-config - **13 accepted after 13A:** trusted verified supported-image admission, explicit new-config
scalar/table projection, equivalent full scenarios/initial state, actual bounded scalar/table projection, equivalent full scenarios/initial state, actual bounded
diagnostic captures and canonical idle/vehicle/fuel/shift comparison with diagnostic captures and canonical idle/vehicle/fuel/shift comparison with
lineage/uncertainty. Save/readback/replay and native demonstration reuse accepted lineage/uncertainty. Save/readback/replay and native demonstration reuse accepted
contracts. Independent tests and supported timeout/identity/invalid/interruption/ contracts. Independent tests and supported timeout/identity/invalid/interruption/
corruption/recovery/config/DTC failure coverage retain earlier regressions; corruption/recovery/config/DTC failure coverage retain earlier regressions;
no invented firmware checksum/partial physical-write protocol. no invented firmware checksum/partial physical-write protocol.
- **14:** verify OBDX Pro GT protocols, drivers, OS/architecture constraints, - **14 readiness assigned:** verify OBDX Pro GT protocols, drivers, OS/architecture constraints,
manufacturer libraries/SDKs, actual J2534 support and C# integration using manufacturer libraries/SDKs, actual J2534 support and C# integration using
attributable sources before implementation. Missing device/driver access means attributable sources before implementation. Current assignment is sourced
real-hardware verification is blocked, not passed by mocks. Restrict services to readiness/specification only: no adapter code, driver install, device open or
verified read-only allowlists; diagnostic access does not imply every module is vehicle traffic. Review the smallest supported read-only implementation next
discoverable or programmable. within the same milestone; missing device access gates real-hardware proof,
not documentation research. No mock-as-hardware or assumed module capability.
### Longer-term scope preservation ### Longer-term scope preservation
@ -1126,14 +1149,21 @@ history, not repeated redesign without evidence.
IsBusy=false before owner Release and drain completion. New admission can IsBusy=false before owner Release and drain completion. New admission can
replace `_drain` while cleanup remains pending. Prepared [Step-13A](Step-13A.md); replace `_drain` while cleanup remains pending. Prepared [Step-13A](Step-13A.md);
no model permissions, hardware scope or unrelated dynamics fixes. no model permissions, hardware scope or unrelated dynamics fixes.
- Accepted Step-13A and corrected Step-13 after reviewed seven actual pre-fix
failures, nine corrective cases, 179 unit + 475 integration = 654 passes,
clean nine-project Release and rebuilt affected native evidence. R-01 closed.
- Strengthened shared worker gate prospectively with source-ordering/cleanup-
tail/final-evidence ledger and prepared [Step-14](Step-14.md) hardware-readiness
research checkpoint. No source/driver/hardware operation or auto handoff.
### Remaining ### Remaining
- Manually execute [Step-13A](Step-13A.md) and shared gate, then review controlled - Manually execute [Step-14](Step-14.md) readiness research and revised shared
actual pre-fix/corrected release-tail/Close/new admission proof and regressions. gate, then review source-backed GT requirements and smallest read-only contract
- Step-14 is blocked/unassigned. Step-13 code and native comparison evidence are before assigning adapter implementation. No Step-15 entry on research alone.
present but not accepted pending C-07 correction. Local-only generation and - Step-13/13A and earlier work are accepted. Step-14 physical integration is
earlier work remain accepted; preserve all numerical/history/safety gates. unimplemented; device/driver/native prerequisites and proof remain gated.
Preserve all numerical/history/safety and manual workflow boundaries.
- Profile/module metadata, synthetic virtual-module communication and deterministic - Profile/module metadata, synthetic virtual-module communication and deterministic
dynamics/v2 and corrected dashboard are accepted; later features remain planned. dynamics/v2 and corrected dashboard are accepted; later features remain planned.
Preserve the complete sequence/safety gates. Preserve the complete sequence/safety gates.
@ -1148,18 +1178,22 @@ Runtime/model availability is an environmental prerequisite, not a permanent
attestation; future missing/changed state must be explicit, with no cloud/model attestation; future missing/changed state must be explicit, with no cloud/model
fallback or automatic setup added to the app. The orchestrator ran no checks. fallback or automatic setup added to the app. The orchestrator ran no checks.
Step-13 R-01 is a concrete new completion-order issue: `_owner=null` and Step-13 R-01 is closed by accepted Step-13A: actual pre-fix tail failures
`IsBusy=false` precede `owner.Release()` and completion of `_drain`. established premature owner/drain replacement, corrected release-before-atomic-
A delayed cancellation callback/release can leave old work undrained while settlement keeps admission reserved through cleanup. Original report remains
new Evaluate/Save replaces the tracked completion. The orchestrator identified unchanged; the added prospective worker ledger prevents relying on pre-work
source ordering, not an independently executed failure; Step-13A requires actual barriers/prose alone for future completion-tail behavior.
controlled reproduction and correction, reusing accepted local-generation
completion prior art without cleanup under required gates. Step-14 exact GT interface/API/device/Windows prerequisites have not been
verified. Research must distinguish supported primary-source facts, observed
local environment and unknowns. No hardware capability or driver/API exists
merely because a product has OBD/J2534/Pro in its name. Current handoff does not
authorize opening a physical adapter or transmitting to a vehicle.
Step-13's actual model/proof/config/metrics/archive/replay evidence remains Step-13's actual model/proof/config/metrics/archive/replay evidence remains
preserved. Fuel changes are synthetic estimates, no physical/checksum claim. preserved. Fuel changes are synthetic estimates, no physical/checksum claim.
Its disclosed intermittent unchanged dynamics regression and known recording Its disclosed intermittent unchanged dynamics regression and known recording
navigation issue are separately tracked, not automatic Step-13A fixes. navigation issue are separately tracked, not automatic Step-14 fixes.
No blocking Step-10 deficiency remains. Its pending-reopen finding below is No blocking Step-10 deficiency remains. Its pending-reopen finding below is
historical and closed by accepted Step-10A: historical and closed by accepted Step-10A:

View File

@ -5,11 +5,11 @@
| Field | Value | | Field | Value |
|---|---| |---|---|
| Updated | 2026-11-01 | | Updated | 2026-11-01 |
| Phase | Step-13 reviewed; comparison ownership/drain tail requires focused correction | | Phase | Corrected Step-13 accepted; bounded hardware-readiness research next |
| Active step | Step-13A - Retain comparison ownership through cleanup and drain completion | | Active step | Step-14 - OBDX Pro GT readiness and read-only integration specification |
| Assignment | [Step-13A.md](Step-13A.md) | | Assignment | [Step-14.md](Step-14.md) |
| Prerequisites | Step-12 with Step-12A ACCEPTED 2026-10-31 | | Prerequisites | Step-13 with Step-13A ACCEPTED 2026-11-01 |
| Current executable handoff | [Step-13A.md](Step-13A.md) - one original C-07 lifetime correction | | Current executable handoff | [Step-14.md](Step-14.md) - research/readiness only, no adapter implementation |
| Assignment status | READY FOR MANUAL WORKER HANDOFF | | Assignment status | READY FOR MANUAL WORKER HANDOFF |
| Completion handoff status | [Step-12-Completion.md](Step-12-Completion.md) superseded for execution; preserve unchanged | | Completion handoff status | [Step-12-Completion.md](Step-12-Completion.md) superseded for execution; preserve unchanged |
| Original worker report | [Step-12-Report.md](Step-12-Report.md) present and reviewed; preserve INCOMPLETE/BLOCKED observations | | Original worker report | [Step-12-Report.md](Step-12-Report.md) present and reviewed; preserve INCOMPLETE/BLOCKED observations |
@ -17,11 +17,12 @@
| Worker report | [Step-12A-Report.md](Step-12A-Report.md) present and reviewed; explicitly INCOMPLETE/BLOCKED | | Worker report | [Step-12A-Report.md](Step-12A-Report.md) present and reviewed; explicitly INCOMPLETE/BLOCKED |
| Accepted execution report | [Step-12A-Execution-Report.md](Step-12A-Execution-Report.md) - reviewed; original blocked reports preserved | | Accepted execution report | [Step-12A-Execution-Report.md](Step-12A-Execution-Report.md) - reviewed; original blocked reports preserved |
| Step-13 worker report | [Step-13-Report.md](Step-13-Report.md) present and reviewed; preserve unchanged | | Step-13 worker report | [Step-13-Report.md](Step-13-Report.md) present and reviewed; preserve unchanged |
| Expected worker report | `prompts/Step-13A-Report.md` - not yet created | | Step-13A worker report | [Step-13A-Report.md](Step-13A-Report.md) present and accepted on reviewed evidence |
| Expected worker report | `prompts/Step-14-Report.md` - not yet created |
| Owner runtime policy | Local models only; no Ollama cloud or cloud fallback, reaffirmed 2026-10-31 | | Owner runtime policy | Local models only; no Ollama cloud or cloud fallback, reaffirmed 2026-10-31 |
| Last accepted correction/follow-up | Step-12A - accepted 2026-10-31; includes evidenced Hybrid dispatch fixes | | Last accepted correction/follow-up | Step-13A - accepted 2026-11-01; R-01 closed |
| Last accepted major milestone | Step-12 with Step-12A - accepted 2026-10-31 | | Last accepted major milestone | Step-13 with Step-13A - accepted 2026-11-01 |
| Outstanding follow-up | Step-13A / R-01 comparison owner release and drain publication; no Step-12B | | Outstanding follow-up | None; no Step-13B required |
| Step-02 status | ACCEPTED after Step-02A | | Step-02 status | ACCEPTED after Step-02A |
| Step-03 status | ACCEPTED after Step-03A/Step-03B | | Step-03 status | ACCEPTED after Step-03A/Step-03B |
| Step-04 status | ACCEPTED after Step-04A/Step-04B | | Step-04 status | ACCEPTED after Step-04A/Step-04B |
@ -37,11 +38,12 @@
| Step-11 status | ACCEPTED 2026-10-31; no correction required | | Step-11 status | ACCEPTED 2026-10-31; no correction required |
| Step-12 status | ACCEPTED 2026-10-31 after Step-12A actual local-model/native qualification | | Step-12 status | ACCEPTED 2026-10-31 after Step-12A actual local-model/native qualification |
| Step-12A status | ACCEPTED 2026-10-31; all frozen setup/model/native/regression gates closed | | Step-12A status | ACCEPTED 2026-10-31; all frozen setup/model/native/regression gates closed |
| Step-13 status | NEEDS CORRECTION for C-07 / AC-03; other implementation/native evidence substantially supported | | Step-13 status | ACCEPTED 2026-11-01 after Step-13A; proof/projection/comparison/native and corrected lifetime supported |
| Step-13A status | READY FOR MANUAL WORKER HANDOFF; no hardware/new feature scope | | Step-13A status | ACCEPTED 2026-11-01; actual seven pre-fix failures and nine corrected cases reviewed |
| Worker execution | Step-13 returned COMPLETE; 179 unit/466 integration = 645 passes and native real-model/target/run/archive evidence reviewed | | Step-14 status | READY FOR MANUAL WORKER HANDOFF; readiness checkpoint only, physical integration remains unimplemented |
| Immediate next action | Manually give worker Step-13A, original Step-13 and shared gate; no automatic invocation | | Worker execution | Step-13A returned COMPLETE; 179 unit/475 integration = 654 passes, clean Release and rebuilt affected native evidence reviewed |
| Next orchestrator action | Review executed pre-fix/corrected ownership-tail cases and regressions before Step-13 acceptance | | Immediate next action | Manually give worker Step-14 and revised shared gate; no automatic invocation or hardware traffic |
| Next orchestrator action | Review sourced hardware-readiness report before assigning any adapter code; no hardware milestone acceptance from research alone |
## Repository baseline at initialization ## Repository baseline at initialization
@ -1651,6 +1653,73 @@ Neither is automatic Step-13A cleanup or a reason to edit accepted dynamics.
not another model-permission gate or new feature. Last accepted major remains not another model-permission gate or new feature. Last accepted major remains
Step-12/12A; Step-14 remains blocked and no hardware assignment is created. Step-12/12A; Step-14 remains blocked and no hardware assignment is created.
## Step-13A acceptance and corrected Step-13
Decision: **ACCEPTED**, 2026-11-01, for Step-13A and corrected Step-13.
R-01 is closed; no Step-13B required. Original Step-13 report and its native
model/proof/projection/capture/replay evidence remain unchanged.
Reviewed the complete [corrective report](Step-13A-Report.md), actual comparison
completion/Close diff, internal observation/test visibility and all new ownership
tests, related docs, frozen assignment hash, unchanged nine-project references
and existing worker Release outputs. The orchestrator ran no application
builds/tests/native/inference/runtime or hardware commands.
| Criterion | Decision | Evidence |
|---|---|---|
| AC-01 | ACCEPTED | Seven executed pre-fix behavioral failures; actual cancellation callback/release held after real work returns, Evaluate/Save Busy with unchanged owner/task/config/target/store effects after correction |
| AC-02 | ACCEPTED | Callback/release outside state gate, actual release still attempted on observation failure; primary/cleanup errors retained; same reserved owner/terminal/exact task settlement and shared Close publication under coherent gates |
| AC-03 | ACCEPTED | Exact old drain stays incomplete/reserved until release; no replacement admission, Close/Dispose awaits it; successful retry only after drain and independent fresh closed/page lifetime |
| AC-04 | ACCEPTED ON WORKER EXECUTION EVIDENCE | Reported 179 unit + 475 integration = 654 passes, zero failed/skipped; all 645 prior cases retained, nine new cases; clean nine-project Release and actual rebuilt manual comparison/save/Busy/Cancel/unmount/native recheck |
| AC-05 | ACCEPTED | Own post-green source path/ledger, truthful internal/native/historical-model distinction, aligned docs and original report preservation/task-only cleanup/manual STOP |
The one internal ReleasingOwner observation is bounded test evidence, not a
public orchestration hook or private model oracle. It neither bypasses the real
workflow nor supplies values; native production has no subscriber.
Owner release now precedes owner removal and exact task settlement. Post-task
notification mutates no successor ownership. Failed cleanup settles waiters as
failure with primary/save evidence retained, not a healthy fallback.
Worker native recheck preserved original/derived fingerprints, 3361 steps/
336.1 s, 1346 actual observations per capture and exact 650/1250 rpm behavior/
unchanged torque axes and unrelated fields. The prior Step-13 real LLM response
and replay proof are historical; no new inference/native replay is claimed.
The prior unchanged dynamics/recording-navigation disclosures remain separate,
not causal corrections; neither recurred in this reported corrective run.
## Prospective worker-control revision and Step-14 checkpoint
The owner requested fewer corrective steps. Accountability is shared: the
orchestrator's earlier gate required pending-work tests but did not explicitly
require the post-work callback/release/drain boundary; a prose self-review left
that gap untested. More assertions/test totals alone would not fix the process.
Revised [WORKER-QUALITY-GATE.md](WORKER-QUALITY-GATE.md) prospectively from
Step-14: inspect accepted owner prior art before adding logic; execute admission,
publication AND completion-tail cases where relevant; include exact source/
invariant/named-evidence ledger; validate the final production edits/rebuilt
affected paths; reconcile every frozen criterion before COMPLETE.
Worker fixes tightly coupled findings inside the active assignment and returns
BLOCKED/INCOMPLETE for genuinely missing evidence. No retrospective criteria
were added to Step-13A and no automatic agent/worker management is introduced.
The orchestrator will use the same frozen rows, consolidate review findings,
and keep new dependency/state-machine prerequisites out of broad assignments.
[Step-14](Step-14.md) is deliberately a read-only research/readiness checkpoint:
sourced exact OBDX Pro GT interface/protocol/Windows/library/support matrix,
read-only local prerequisite inventory if accessible, and a bounded initial
adapter contract/qualification specification. No application/adapter source,
driver install, device opening or vehicle traffic is authorized in this prompt.
It must not claim physical integration or hardware acceptance from documentation,
enumeration or mocks. Missing device access is an explicit later integration
gate, not a reason to fabricate research conclusions.
After this checkpoint returns, review the sourced specification and dependencies
before authorizing the smallest actual read-only implementation as a separate
manual handoff within the Step-14 milestone. Do not auto-advance Step-15 or
renumber completed history. Current hardware behavior remains unimplemented.
No further model/runtime changes or unrelated baseline repairs are assigned.
## Prerequisites, limitations, and known issues ## Prerequisites, limitations, and known issues
- Step-01A worker evidence establishes portable and native Windows compatibility - Step-01A worker evidence establishes portable and native Windows compatibility
@ -1689,22 +1758,19 @@ Step-12/12A; Step-14 remains blocked and no hardware assignment is created.
## Review and progression gate ## Review and progression gate
On manual return, evaluate evidence from [Step-13A](Step-13A.md) On manual return, evaluate [Step-14](Step-14.md) readiness evidence against its
against its matrix, original Step-13 C-07/AC-03 and shared gate using both reports, frozen research matrix and revised shared gate, cited primary sources, exact
actual code/tests, dependency graph, Git changes, device/interface assertions, local prerequisite observations and proposed
README/docs, and recorded build/test evidence. Do not accept solely on a success bounded adapter/test contract. No application execution is assigned.
claim; do not claim independent runtime verification.
- Inspect actual pre-fix release-tail/competing admission/Close drain cases, - Inspect attributable exact GT evidence, version/architecture/protocol/API/
atomic corrected completion, failure/result/save evidence and existing native licensing limits, unknowns and supported initial read-only boundary. Do not
comparison behavior. Preserve supported projection/run/model contracts. infer capability from a product name, generic OBD family or synthetic v1/v2.
- If accepted: accept Step-13A and corrected Step-13, update roadmap/state, prepare bounded Step-14 read-only hardware work only with its prerequisites - If readiness accepted: record the checkpoint only and prepare one narrow
and stop for manual handoff. actual integration handoff within Step-14 after prerequisite review; do not
- If deficient: retain Step-13A open and identify a material unmet contract/ accept the full hardware milestone or advance Step-15 yet.
regression; any further corrective assignment must be narrowly evidenced. - If deficient/blocked: identify the exact unsupported source/contract/prerequisite
Planned new features are not corrections. and retain the checkpoint; no guessed implementation or automatic new letter.
- If blocked: retain the active step, document the prerequisite and required
worker evidence; do not advance or mark unexecuted checks as successful.
- Never create the next major assignment while acceptance or corrections are open. - Never create the next major assignment while acceptance or corrections are open.
## History ## History
@ -1798,12 +1864,15 @@ claim; do not claim independent runtime verification.
| 2026-10-31 | Step-13 prepared | Explicit verified supported-image projection to NEW immutable configs, equivalent real scenario/capture/comparison/replay and native demonstration; original source/rehearsal isolation preserved, no automatic handoff | | 2026-10-31 | Step-13 prepared | Explicit verified supported-image projection to NEW immutable configs, equivalent real scenario/capture/comparison/replay and native demonstration; original source/rehearsal isolation preserved, no automatic handoff |
| 2026-11-01 | Step-13 returned and reviewed | Reported 645-case/clean Release and actual model/proof/new-config/real-run/archive/replay evidence inspected; most frozen requirements supported | | 2026-11-01 | Step-13 returned and reviewed | Reported 645-case/clean Release and actual model/proof/new-config/real-run/archive/replay evidence inspected; most frozen requirements supported |
| 2026-11-01 | Step-13 needs focused C-07 correction; Step-13A prepared | Owner/IsBusy cleared before owner Release and public drain settlement permits replacement of undrained completion; source ordering finding, controlled pre-fix worker proof required; no Step-14 | | 2026-11-01 | Step-13 needs focused C-07 correction; Step-13A prepared | Owner/IsBusy cleared before owner Release and public drain settlement permits replacement of undrained completion; source ordering finding, controlled pre-fix worker proof required; no Step-14 |
| 2026-11-01 | Step-13A returned and ACCEPTED; corrected Step-13 ACCEPTED | Actual seven pre-fix tail failures, nine corrected cases/654 passes/clean Release/rebuilt native source evidence reviewed; ownership reserved through release and atomic exact drain settlement, no 13B |
| 2026-11-01 | Worker control strengthened prospectively; Step-14 readiness prepared | Required boundary ledger/completion-tail/final changed-path evidence from Step-14, tighter prerequisite-first scope; manual sourced hardware readiness only, no adapter code/vehicle traffic or retrospective gate |
## Manual handoff instruction ## Manual handoff instruction
Provide [Step-13A.md](Step-13A.md), original [Step-13.md](Step-13.md) and shared Provide [Step-14.md](Step-14.md) and revised [worker gate](WORKER-QUALITY-GATE.md)
[worker gate](WORKER-QUALITY-GATE.md) to the separate worker. Execute Step-13A to the separate worker. Execute the readiness checkpoint only and return
only and return `prompts/Step-13A-Report.md`, preserving the Step-13 report and `prompts/Step-14-Report.md` plus its requested sourced readiness document.
all earlier assignments/reports. Step-13 remains unaccepted; no Step-14. No application code/driver install/hardware traffic. Stop for review before
any implementation handoff; full Step-14 hardware acceptance remains pending.
Return manually to the orchestrator for review. Reports, including final chat Return manually to the orchestrator for review. Reports, including final chat
reports, must follow the single plain triple-backtick copy-window requirement. reports, must follow the single plain triple-backtick copy-window requirement.

328
prompts/Step-13A-Report.md Normal file
View File

@ -0,0 +1,328 @@
```
STEP: Step-13A - Retain comparison ownership through cleanup and drain completion
STATUS: COMPLETE - R-01 corrective worker execution verified.
MANUAL GATE: Step-13/13A acceptance remains pending; STOP. Step-14 unassigned.
DATE: 2026-11-01
BASE: 18c9d0d ("WIP 13"); initial worktree clean; no staging/commit/push.
R-01 AND ACTUAL PRE-FIX REPRODUCTION
The previous ObserveAsync cleared _owner and published terminal IsBusy=false
before owner.Release and before settling the exact task stored in _drain.
A cancellation callback holding the owner's private cancellation gate could
leave release blocked while a new Evaluate/Save replaced _drain.
Close could then follow only the newer completion and miss the original owner.
Added and EXECUTED the real feature-boundary regressions BEFORE changing this
production order. The only initial production additions were an internal release
observation event and test-assembly visibility; neither fixed/reordered admission,
release or terminal publication. Real owning rehearsal/projection/dynamics runs
and actual encoded captures prepared the comparison.
Pre-fix command:
dotnet test .\src\tests\AutoForge.IntegrationTests\AutoForge.IntegrationTests.csproj --configuration Release --no-restore --filter "FullyQualifiedName~ComparisonOwnershipTail_" --verbosity minimal
FAILED; exit 1; 7 failed, 0 passed, 0 skipped; approximately 6 s.
Compilation succeeded. All barriers/resources were released/drained in finally.
This is a behavioral reproduction, not a prose race or compilation failure.
| Actual pre-fix condition | Observed failure |
|---|---|
| Successful real pair Save; callback/release held; competing Save | tail busy=False; state=Completed; old drain completed=False; callback exited=False; save admitted=True |
| Successful real pair Save; callback/release held; competing Evaluate | tail busy=False; state=Completed; old drain completed=False; callback exited=False; evaluate admitted=True |
| Actual final-readback failure; callback/release held; competing Save | tail busy=False; state=Failed; old drain completed=False; callback exited=False; save admitted=True |
| Actual final-readback failure; callback/release held; competing Evaluate | tail busy=False; state=Failed; old drain completed=False; callback exited=False; evaluate admitted=True |
| New Save replaced drain; Close | replacement admitted=True; old drain completed=False; callback exited=False; Close/Dispose completed=True; state=Closed |
| New Save replaced drain; Dispose | replacement admitted=True; old drain completed=False; callback exited=False; Close/Dispose completed=True; state=Closed |
| Actual cancelled evaluation; completion-release boundary held | cancelled evaluation tail busy=False; state=Cancelled; old drain completed=False |
CONTROLLED BOUNDARY AND NARROW TEST SEAM
CompletionTailStore wraps the accepted real comparison capture/codec/store data.
It registers a callback on the actual owner token passed through SaveAsync.
The callback reads the real comparison snapshot, then remains synchronously
held. CancelAsync is invoked on a test worker because this actual callback holds
Owner.Cancel's private gate. Final archive readback can then return or fail, so
operation work finishes while actual owner.Release remains blocked.
No comparison coordinator, physics, signal samples, target proof or saved result
is mocked. The successful delayed readback is the actual already persisted
capture, not a cancellation-to-success fallback invented by the coordinator.
Existing Changed notification occurs only after completion; it cannot signal
this post-work/pre-release interval without timing polling. Therefore the one
minimal INTERNAL ReleasingOwner(Task) observation event signals the actual
release boundary and exact original completion task. It supplies no public
controls, values, proof, metrics or fake successful completion. The cancelled
evaluation case also holds this real release boundary after actual signal
cancellation/diagnostic disposal; cleanup-error cases exercise observer failure.
Application grants only IntegrationTests internal visibility, following existing
test-friend prior art. Native production has no subscriber to this event.
No sleeps/polling, alternate host or new lifetime framework.
IMPLEMENTATION AND ATOMIC ORDER
1. Start reserves the real owner, generation and exact completion/_drain under
the comparison gate, with Evaluating/Saving and IsBusy=true.
2. Actual operation work finishes or faults/cancels. Ownership stays reserved.
3. Release observation and owner.Release execute OUTSIDE the comparison gate.
A held callback/release cannot expose idle or admit another operation.
4. Observation/release errors are explicitly captured. Actual release is still
attempted after observation failure. Primary and additional errors are
retained together with AggregateException rather than masking either.
5. Under ONE comparison gate, remove the same reserved owner, publish truthful
Completed/Failed/Cancelled state (unless Closing owns terminal publication),
and settle that exact operation completion task. Continuations are async.
6. Changed notification follows outside the gate. No old owner can clear or
overwrite a legitimately admitted successor because admission was blocked
until release and completion settlement finished.
7. Close retains its shared task and captured owner/drain, cancels and awaits
outside required gates, then publishes Closed/IsBusy=false and settles its
shared completion under the same gate. Existing visible Close error and
result-retirement semantics are preserved.
No callback/release/await was moved under the comparison state gate.
The owner remains the reservation, not a second duplicate owner framework.
The specific old _drain cannot be replaced during its unfinished tail.
FILES AND PUBLIC CONTRACTS
| File | Change |
|---|---|
| [CalibrationComparisonSession.cs](../src/AutoForge.Application/Features/Vehicles/Calibration/CalibrationComparisonSession.cs) | Release before atomic owner/terminal/completion publication; explicit cleanup error capture/aggregation; atomic shared Close terminal settlement; one internal real-release observation event |
| [AssemblyInfo.cs](../src/AutoForge.Application/Properties/AssemblyInfo.cs) | IntegrationTests-only internal visibility for that observation seam |
| [CalibrationComparisonOwnershipTests.cs](../src/tests/AutoForge.IntegrationTests/CalibrationComparisonOwnershipTests.cs) | Nine real ownership-tail/cleanup cases reusing existing owning comparison fixtures, actual models/transport/captures and store codec |
| [README.md](../README.md) | Corrective ownership-tail status and separate report link; manual review still pending |
| [documentation index](../docs/README.md) | Step-13/13A pending acceptance status |
| [comparison usage](../docs/calibration-comparison.md) | Busy through release, exact drain/Close ordering, primary/cleanup error and internal-vs-native evidence semantics |
| [architecture](../docs/architecture.md) | Actual reservation/release/publication gate order without new graph or framework |
| [build/test](../docs/build-and-test.md) | Separate corrective pre-fix/focused/full/native evidence; preserved historical Step-13/LLM claims |
| [this report](Step-13A-Report.md) | Sole new file under prompts; original reports/assignments/state unchanged |
Public comparison/source/projection/archive interfaces are unchanged.
No UI modification, new project/package/reference/host/schema/protocol/physics/
metrics/proof change. Core/Simulation/Diagnostics and independent Components
remain unchanged; Application/UI gain no private numerical oracle.
FROZEN CORRECTIVE MATRIX - NAMED EXECUTED EVIDENCE
All new named cases are in the ownership test file linked above.
| Contract | Executed evidence and corrected decisive result | Status |
|---|---|---|
| C-01 | ComparisonOwnershipTail_RealCancellationCallbackReservesAdmissionUntilRelease, success/failure x Evaluate/Save = 4 cases. Tail remains Saving/IsBusy=true, original task incomplete and callback unexited; both contenders reject typed Busy before work. Target reads/writes, configuration count and store calls unchanged; original result/save evidence retained. | PASSED |
| C-02 | ComparisonOwnershipTail_CloseOrDisposeIncludesOldDrainAfterCompetingSave = 2 cases. Replacement Save rejects Busy. Close/Dispose includes the old drain and actual callback; Closed notification records old task settled AND callback exited. Callback runs once; retained result Stale; shared Close task retained; closed Save rejects; independent fresh comparison succeeds. | PASSED |
| C-03 | Four success/final-readback-failure cases above, actual cancelled evaluation case below, and ComparisonOwnerCleanup_FailureSettlesWaitersAndPreservesPrimaryAndSaveEvidence = 2 cases. Successful/failed/cancelled terminal publication follows release. Cleanup-only failure returns its IOException; primary Save failure plus cleanup failure preserves both in AggregateException. Waiters settle as failures, not timeout/false success; per-ID publication/readback/error evidence remains. | PASSED |
| C-04 | ComparisonOwnershipTail_CancelledEvaluationKeepsExactDrainUntilCleanupAndThenRetries = 1 case. Actual cancelled signal run disposed once; exact same evaluation drain retained through held cleanup and repeated Cancel; Save rejects Busy; then Cancelled terminal state, bounded partial data and valid next real full evaluation. Other tail cases retry Save only after full release, skip already verified files, and verify fresh closed/page lifetimes. | PASSED |
| C-05 | Combined 59-case selector, full 179 unit + 475 integration = 654, actual nine-project Release and native affected commands. All 645 prior cases retained; unchanged metrics/fingerprints/caps/proof/authority and old manual/unit/source behavior verified. | PASSED |
| Acceptance | Mapping and executed criterion | Status |
|---|---|---|
| AC-01 | Actual 7-case behavioral pre-fix failure; corrected C-01 rejects both contenders with zero new work effects while old release is held | PASSED |
| AC-02 | C-02/C-03 real callback/drain/Close/Dispose, once-only cleanup, truthful waiter and primary/partial-save error outcomes | PASSED |
| AC-03 | C-04 same old completion reservation, no premature/late successor mutation, valid post-drain retry and fresh independent native lifetime | PASSED |
| AC-04 | C-05 645 retained plus 9 new cases, nine-project Release, actual affected native comparison/save/report/owner checks | PASSED |
| AC-05 | Post-green source-path self-review, directly related docs, protected history/cleanup and separate single fenced report | PASSED; manual acceptance pending |
CORRECTED COMMANDS AND COUNTS
Test platform: VSTest / xUnit v2, no MTP separator/flags.
Current SDK dotnet --version: 10.0.403; permitted latestPatch roll-forward from
the existing global.json 10.0.402. No SDK/workload/package/runtime update.
Focused combined command:
dotnet test .\src\tests\AutoForge.IntegrationTests\AutoForge.IntegrationTests.csproj --configuration Release --no-restore --filter "FullyQualifiedName~ComparisonOwnershipTail_|FullyQualifiedName~ComparisonOwnerCleanup_|FullyQualifiedName~Comparison_|FullyQualifiedName~PreviewHandoff_|FullyQualifiedName~TorqueDraft_" --verbosity minimal
PASSED; exit 0; 59 passed, 0 failed, 0 skipped; approximately 12 s.
Nine new corrective cases plus the retained 50-case comparison/pinned-draft group.
Required sequential commands:
dotnet build .\src\AutoForge.slnx --configuration Release --no-restore
PASSED; exit 0; all nine projects; 0 warnings, 0 errors; 5.80 s.
dotnet test .\src\AutoForge.slnx --configuration Release --no-build
PASSED; exit 0; 179 unit + 475 integration = 654; 0 failed, 0 skipped.
Unit approximately 238 ms; integration approximately 12 s.
All original 645 cases preserved, plus 9 corrective integration cases.
| Project | Build | Tests |
|---|---|---|
| AutoForge.Core | PASSED | N/A, non-test |
| AutoForge.Diagnostics | PASSED | N/A, non-test |
| AutoForge.Simulation | PASSED | N/A, non-test |
| AutoForge.Application | PASSED | N/A, non-test |
| AutoForge.Components | PASSED | N/A, non-test |
| AutoForge.UI | PASSED | N/A, non-test |
| AutoForge.Client | PASSED, actual Windows MAUI Hybrid executable | N/A |
| AutoForge.UnitTests | PASSED | 179 passed, 0 failed/skipped |
| AutoForge.IntegrationTests | PASSED | 475 passed, 0 failed/skipped |
Prior Step-13's intermittent unchanged dynamics TaskCanceledException and
recording-navigation Busy issue remain separate historical disclosures.
Neither occurred in this corrective full/native execution. Their source files
have no diff; no causality inferred from the prior flaky run and no unrelated
repair/waiver. No dependencies restored/installed because none were missing.
ACTUAL REBUILT NATIVE CHECKS - NOT CONTROLLED TAIL OR NEW LLM CLAIMS
Actual [Client executable](../src/AutoForge.Client/bin/Release/net10.0-windows10.0.26100.0/win-x64/AutoForge.Client.exe);
task PID 40220, created 2026-11-01T09:06:02.1060905-05:00.
Task-owned isolated step13a-webview/CDP 9387, not an alternate app host.
Application assembly SHA256:
561A6E48E4C2F5AB441FA5202ADCBCB3F0C05D53D3C2D5FC692FA217FFE5A58B.
Initial user preferences Metric/Road; initial archive files 0.
No test release subscriber exists in native production.
Explicit actual synthetic profile/revision -> warm original -> Baseline ->
MANUAL runbook template/validation -> pinned complete draft -> separate Save ->
fresh isolated target/two-read backup/prepare/confirmation/approval/Apply/full
actual readback -> owning comparison -> explicit two-capture Save/readback.
No repeated inference was run. Original real LLM content and qualification in
the preserved [Step-13 report](Step-13-Report.md) remain HISTORICAL evidence.
No model/runtime permission forms, chat, pull/install/restart/config/unload.
Original configuration:
sha256:v1:e8991ae139b54959ec53a4fe64b08dc9bfc03feb533496cf741f2c9f3a2342da
Derived NEW configuration:
sha256:v1:3ee252697bfb69f36f4b99dca4c0267f7517a17a459666fb47551fa6ed978450
Workspace 7ce9775e197043fc9c8da2d3f8d5f98e.
Baseline a54ea42c082d4d56891e5b3c9c02e4c7, revision 1.
Proposed 00eabb196e53472cb545c21a712696ed, revision 2; history 2/32.
Copy alone retained revision 1; separate Save created revision 2.
Exact scalar 650->1250 rpm; cells [240,300,360,350,300,180] ->
[240,300,350,350,300,180] N*m; axis [650,1000,2000,3200,4200,5000] unchanged.
Actual target instance 1132fcd0d939427f89d6e16334496dcb, revision 0->1.
Two-read backup 195a79aa86864c6bab1ec437b010b575.
Preparation a995b7be3ea64678bda3af36efe20b46.
Explicit approval 915e552f3e5947b7bb05d3b9ab54df16.
Actual Apply 47ee64ca9a4d4b7e851e95a9bea48959;
completed 2026-11-01T14:07:14.9347123+00:00.
Acknowledgement True/revision 1, separately verified actual complete image.
VerifiedApplied / CandidateMatching with every unchanged cell/axis/unit checked.
Native comparison e3aa975405424ba2a0f6dbe5a02239c0;
proof a8084d560cd64ea588ce25de5e7ca6a8;
proof as of 2026-11-01T14:07:15.0903722+00:00.
Profile autoforge.synthetic.test-truck@1; engine module.engine.primary;
transmission module.transmission.primary; model autoforge.synthetic.longitudinal@1;
definition autoforge.synthetic.engine-authoring@1.
Provider autoforge.simulation.dynamics, actual v2 signals.
| Real native metric | Baseline | Derived |
|---|---|---|
| Run/session | dynamics-run-000001 / diagnostic-session-000001 | dynamics-run-000002 / diagnostic-session-000002 |
| Capture | bf2305069faa4c8fb2ba3fb3404a7cfc | 0905ad908ef741058e17ca6dd69d5ba9 |
| Completion / actual observations | Complete / 1346 | Complete / 1346 |
| Warm target rpm | 650 | 1250 |
| Actual last-ten maximum idle error rpm | 4.084768079337664E-10 | 7.855760486563668E-10 |
| Actual last-ten maximum stationary speed m/s | 0 | 0 |
| Model steps / total virtual seconds | 3361 / 336.1 | 3361 / 336.1 |
| Warm/Accelerate/Cruise/Brake/ReturnIdle steps | 100/103/3000/58/100 | 100/103/3000/58/100 |
| Acceleration / brake seconds | 10.3 / 5.800000000000001 | 10.3 / 5.800000000000001 |
| Sampled cruise count / speed range m/s | 600 / 20.1168..20.1168 | 600 / 20.1168..20.1168 |
| Fuel consumed kg | 0.518385971679477 | 0.5184897031483956 |
| Distance m | 6193.507764002321 | 6193.389583605566 |
| Final actual shift count | 4 | 4 |
Data/proof/fingerprints remain the supported Step-13 semantics, not redesigned.
3900 steps/1562 scheduled observations, both modules every five steps plus
completion, and existing 4096-observation/16 MiB observation/1 MiB header/
32 MiB archive bounds unchanged. Sparse ranges are not continuous extrema;
final shift count is not a complete shift timeline or real-world improvement.
Native Save completed with BOTH persisted True/readback verified True and
"exact readback verified" per ID. Independent saved-byte inspection also
validated actual final raw response step 3361/time 336.1, matching exact
header config/run and transmission.shift-count 4/count/Good.
Baseline file 8670832 bytes; encoded observations 8668539 bytes; SHA256
ed5560f957deb38b06f439b3f7cab014d5dd2c75f0b71ebbcb2fc6241de3e902.
Derived file 8668935 bytes; encoded observations 8666642 bytes; SHA256
ea5d6e02b06376ed93bb3d03ad2adcf3ed8f1af278a704554a341674338f3797.
No overwrite, compensating deletion or altered source/recording owner.
Affected native owner checks:
- Observed "Evaluating; operation/drain pending True." BEFORE competing action/
Cancel, not a completed request retrospectively labelled pending.
- Competing Evaluate visibly rejected "One comparison/run/save owner is active;
cancel and drain first."
- Cancel reached "Cancelled; operation/drain pending False.", visible cancellation,
same last successful report retained Stale/as-of. Save evidence preserved.
- Global ImperialUS toggle/restored Metric/Road left canonical report exactly
unchanged except the legitimate Current->Stale retirement label.
- Close reached Closed; new Evaluate rejected "This page comparison owner is closed."
- Current-head manual template validation still produced Current / NoChanges.
- Reentry created Idle/fresh owner with no old comparison report/authority and
unchanged authoring history 2/32.
- Fresh target was explicitly backed up/approved/applied again; actual new Apply
7baee970d538423fabb73bfad3b6e798. Fresh evaluation unmounted only AFTER observed
pending True; Home "AutoForge foundation" rendered with no late report.
Subsequent calibration entry again created independent Idle owner.
- Final actual original export remained 650/[240,300,360,350,300,180], exact original
five-phase scenario (10/60/300/60/10 s; driving target 20.1168 m/s).
Original Metric/Road preferences restored; Blazor error display none.
These are ordinary actual native affected-lifecycle checks. The held cancellation/
release tail and cleanup-error interleavings are CONTROLLED real-boundary tests,
not claimed to be naturally reproduced native timing or newly generated LLM data.
Existing replay/store boundary tests ran in the 59-case/full selections; no new
native replay/model/physical capability is claimed.
POST-GREEN SOURCE-PATH SELF-REVIEW
Reviewed Start/Evaluate/Save -> actual work/capture/target/store -> ObserveAsync ->
release observation/Owner.Release -> exact task settlement -> Cancel/Close/
Dispose -> UI observer/fresh lifetime. Reused the accepted local-generation
completion order without a general framework.
Confirmed:
- No owner/busy release before cleanup, no _drain replacement interval.
- Callback and release work outside required state gates; actual callback reads
real state without deadlock while holding its own cancellation gate.
- Primary success/error/cancel and per-ID saved/readback/error outcomes preserved.
- Cleanup failure cannot prevent actual release attempt or strand completion;
combined failures retain both, never a success-shaped fallback.
- Terminal owner/state/completion and shared Close terminal task coherent under
the comparison gate; async continuations cannot observe partial publication.
- Old completion has no later state mutation capable of clearing a successor.
- Busy/Closed rejection precedes new work, config/target/capture/store effects;
full-drain retry and independent new lifetime work normally.
- No proof/value/physics/archive checks weakened, no unassigned runtime or file
rollback, and no unrelated baseline failures repaired.
PRESERVATION, DOCUMENTATION AND EXACT CLEANUP
All old prompt reports and assignments preserved; sole new prompt file is this
report. Protected CURRENT hashes rechecked:
[Step-13A.md](Step-13A.md)
DC55E75163EAC8D5A9B4A384AA961C05DCB3087BC7250EC86E7C8511629C7BFA
[STATE.md](STATE.md)
6E912BAEAD65CC36D73BC5A2280178DEAD47DBB55359218F36531C48C6D1EDA3
[ROADMAP.md](ROADMAP.md)
31B4277C62111B9BF86AA46BD1E351931D98DF259A874EBA1D163A701BC534EF
[WORKER-QUALITY-GATE.md](WORKER-QUALITY-GATE.md)
F9C0B14516B70345E4CE93AD1EDB8856094BC92E26A37AEB16EDE687AD021804
[Step-13-Report.md](Step-13-Report.md)
C1D8E1F33D26675189DDC8E559620EAEE7FB342ECB5DFE905C6C2ACF08896340
[Step-13.md](Step-13.md)
121A97B67A6604BE53E8C5742A7F2001971A99F3B6E9E861671C39704B6BA535
[Step-12A-Execution-Report.md](Step-12A-Execution-Report.md)
C5056359F33CC95C4D53F2ACAF7DBAB9E59B8B99CE66B0CD99B3475B427621F5
Git whitespace/scope checks passed. Git LF/CRLF notices are not build warnings.
Nine-project graph/manifests/dependencies unchanged; only internal test visibility.
No agents, permission/model/runtime work, staging/commits/push or Step-14.
EXECUTED cleanup:
- Stopped only freshly path/creation-time-verified task Client PID 40220 and any
exact task-profile WebView descendants; no name-based process termination.
- Removed only the two exact task-generated/verified archive IDs above after
successful Save/readback/raw-context/hash checks. Initial/final user archive
count 0; no unknown/user file or compensating rollback deletion.
- Removed ten exact task native helper/evidence files, inspected step13a-webview
profile, and the exact owned oversized pre-fix temporary output.
- Session root/files directory and all other session data preserved.
Remaining Step13A native targets 0; CDP 9387 listeners 0.
- Restored original Metric/Road preferences.
- Shared Ollama PID 50460, binary and original 2026-10-31 14:39:16 creation/
installed-model/store/configuration remain untouched. No inference rerun.
Persistent result: focused R-01 source/tests/docs and this separate report.
STOP FOR MANUAL ORCHESTRATOR REVIEW; no acceptance or next-step invocation.
```

194
prompts/Step-14.md Normal file
View File

@ -0,0 +1,194 @@
# Step-14 - OBDX Pro GT readiness and read-only integration specification
## STEP IDENTIFIER
Step-14, hardware-readiness checkpoint only.
Step-13 with Step-13A accepted 2026-11-01; 654 cases reported passing.
This is the sole active manual worker assignment. It does not authorize
adapter implementation or accept the full Step-14 hardware milestone.
Read [STATE.md](STATE.md), [ROADMAP.md](ROADMAP.md) and the revised
[WORKER-QUALITY-GATE.md](WORKER-QUALITY-GATE.md). STOP after the requested report.
## STEP TITLE
Establish exact sourced OBDX Pro GT prerequisites and the smallest supported
read-only adapter boundary before any hardware code or vehicle traffic.
## OBJECTIVE
One deliverable: an attributable, reviewable GT readiness specification with
known facts, unresolved prerequisites and a bounded first implementation/test
contract. Do not combine research with a new transport/state machine/UI.
This prerequisite-first scope and final evidence reconciliation are deliberate
controls to reduce preventable corrective handoffs.
Produce `docs/obdx-pro-gt-readiness.md` and `prompts/Step-14-Report.md`.
No application source/tests/project/package/driver/runtime changes, builds/
tests/inference, adapter opening or physical diagnostic transmission.
## ARCHITECTURAL CONTEXT
Existing nine-project .NET 10 modular monolith: Core neutral contracts,
Diagnostics ports/protocol/result validation, Simulation isolated virtual
providers, Application use cases, UI presentation, Client thin Windows MAUI
composition, independent Components and tests under `src/tests`.
Simulation/RecordedReplay/PhysicalHardware are explicit; never substitute one
for another. Existing synthetic JSON v1/v2 is NOT an automotive wire protocol.
Physical read-only support cannot imply programming/clearing/actuation authority.
Inspect the actual [provider contracts](../src/AutoForge.Diagnostics/ProviderContracts.cs),
relevant session/target composition and existing typed result/capture boundaries
only as needed to propose the smallest safe integration. Do not refactor them.
## CURRENT IMPLEMENTATION STATE
- Corrected synthetic comparison/demo accepted; 179 unit + 475 integration =
654 cases, clean nine-project Release and native affected paths reported.
- Real local-only model and synthetic diagnostics/dynamics/proposals/rehearsal/
archive/replay are accepted. They are not evidence of physical support.
- No OBDX Pro GT driver/API/protocol/device qualification is established.
Actual local device ownership, hardware/firmware revision, transport and
approved vehicle/bench access may be unavailable.
- Keep existing source, original assignments/reports, user data and model
runtime unchanged. No new permission form for unrelated model/runtime work.
## REQUIRED IMPLEMENTATION
### Frozen research matrix
| ID | Research action | Required outcome |
|---|---|---|
| C-01 | Verify exact GT product/version/interface and primary documentation | Cite precise primary URLs/revisions/retrieval date; do not borrow another OBDX model's claims without explicit GT applicability |
| C-02 | Establish host/API/protocol/licensing requirements | Distinguish documented native interface, any actual J2534 support, USB/serial/wireless transports, Windows/architecture/driver/library constraints and unsupported/unknown items |
| C-03 | Read-only accessible local prerequisite inventory | Distinguish OS-enumerated device/installed files from an actual tested connection; absent/uncertain access is explicit, no driver install/device open/vehicle traffic |
| C-04 | Specify one smallest evidenced read-only adapter slice | Exact request/reply/units/identity/error/limits/lifecycle contract and real hardware qualification plan, reused project/ports; no unsupported frame/API invention or broad feature backlog |
| C-05 | Review source assertions, negative/lifecycle spec and report completeness | Every factual claim attributable or labelled unknown; future admission/publication/completion-tail tests concrete; no implementation/hardware success claimed from research |
### 1. Exact sourced support matrix
Prioritize manufacturer product/manual/protocol/SDK/driver/license sources.
Use maintained authoritative repositories if linked or verifiably applicable.
Third-party claims are secondary, labelled and cross-checked. Record source
URL/title/version/date and which exact matrix rows each supports.
Identify product variants/firmware/hardware revisions and:
- Documented connection interfaces and host wire framing/API command set.
- Supported vehicle protocols/buses and their exact GT/version conditions.
- Actual J2534 support, if evidenced; do not assume it from a family/name.
- Windows version/CPU architecture, native library ABI/driver/install/access
prerequisites, maintained .NET integration options and distribution/license
constraints. A vendor SDK being downloadable does not prove usability.
- Device metadata versus vehicle/module metadata, supported read-only services
and clear boundaries against writes/reset/clear/actuation/security/programming.
- Timing, framing, request IDs, buffer limits and documented failures relevant
to the first proposed slice. Unknown field sizes are unknown, not invented.
If sources disagree/unavailable, name the uncertainty and evidence needed.
Do not repeatedly search indefinitely or turn missing specs into plausible
implementation instructions. No private code/data/credentials upload.
### 2. Read-only local readiness
If accessible, inspect existing OS device enumeration and installed driver/
library metadata by exact scoped identity/path. Do not scan personal files,
dump environment/serial identifiers or open serial/Bluetooth/device handles.
No pairing, driver/SDK installation, probing hardware commands or vehicle
connection/startup. Public documentation retrieval is permitted.
Record observed/not observed/unverified precisely. Discovery in Device Manager
is not verified physical support. If device access is absent, finish source
research and make the future hardware-test prerequisite explicit.
Do not block all research just because no device can be tested in this task.
### 3. Smallest next implementation contract
Propose ONE primary new hardware contract, preferably explicit adapter
connection/device identification plus one well-documented read-only capability
if the evidence supports it. Do not bundle a complete CAN/UDS/J2534 stack,
multiple adapters/protocols, DTC clearing, dashboards or physical programming.
If no safe actionable API is evidenced, return no-go/required evidence rather
than scaffolding empty interfaces.
Specify existing project/port reuse and necessary minimal additions, without
writing code or new assignment files. The plan must state:
- Exact physical selection/provider identity and no synthetic/replay fallback.
- Eligibility/read-only allowlist before dispatch, actual framing/API outcome
validation before positive result and metadata provenance/unknown handling.
- Bounds/cancellation/deadline/single-owner rules and synchronous-prefix/
post-work release/drain/Close ordering.
- Normal/rejected/boundary/disconnect/overlap cases with concrete test stimulus
and measurable expected output/unchanged state. Use the revised shared
source-order ledger as the future implementation acceptance design.
- Fault injection below actual adapter boundary; simulated/recorded fixtures
labelled controlled, actual installed-device/native qualification separate.
- Specific missing driver/device/authorized bench/vehicle prerequisites and
how real evidence will establish or reject the proposed supported subset.
No hardware safe-to-write claim or generic raw-command UI. Restrict physical
testing in the proposed later plan to approved, documented read-only operations.
The orchestrator will review this before authorizing an implementation handoff.
## FILES OR PROJECTS EXPECTED TO CHANGE
Create the sourced readiness document and sole worker report named above.
Update `docs/README.md` with its link/current research-only status if appropriate.
No other docs need historical rewrite; no source/tests/dependencies/config/
orchestration/gate/old report changes.
## INTERFACE AND DEPENDENCY REQUIREMENTS
Research/specification only. Preserve nine projects/graph, physical restrictions,
source/unit/protocol/archive and owner contracts. No new package/host/runtime/
driver/SDK installation, simulation rewrite, cloud/model operations or firmware
reverse engineering without separate scope. No agents, automatic handoff,
staging/commit/push or shared-server/device changes.
## TESTING REQUIREMENTS
No application build/test or native hardware smoke is assigned: no code changes.
Validate requested document links, source applicability and required output
shape; inspect cited primary evidence instead of trusting search summaries.
Every C-01..C-05 row needs a specific source/observation/specification result.
Source-ordering ledger may be N/A for executed code because none is changed;
its relevant future tests must be concrete in the specification.
Do not manufacture a passing hardware test or rerun 654 cases as a proxy for
research accuracy. Preserve all earlier execution evidence unchanged.
## DOCUMENTATION REQUIREMENTS
The readiness document contains exact GT sourced support matrix, local observed/
missing prerequisites, bounded proposed adapter/test boundary, licensing risks,
unknowns and explicit go/no-go for that subset. Distinguish research readiness
from implemented/tested physical capability.
No standalone planning/checklist files beyond these requested deliverables.
## ACCEPTANCE CRITERIA
| ID | Readiness criterion | Evidence |
|---|---|---|
| AC-01 | C-01/C-02 exact attributable GT interface/API/host/protocol/licensing facts | Primary-source rows, explicit variant/revision scope and labelled unknown/conflicting secondary claims |
| AC-02 | C-03 honest local prerequisites, no physical/installation action | Exact scoped metadata observations or explicit unavailable status; no enumeration-as-tested claim |
| AC-03 | C-04 one bounded implementable read-only subset or justified no-go | Reused real ports, concrete framing/result/owner/negative/hardware test specification and dependencies |
| AC-04 | C-05 final source/coverage reconciliation and preservation | Named source mapping, link checks, self-review and no undocumented/code/hardware claims |
| AC-05 | Research-only report/manual gate truthful | Requested deliverables, precise checkpoint status and STOP before implementation |
## OUT-OF-SCOPE WORK
Adapter/application coding/scaffolding, builds/tests, driver/SDK installation,
device open/pairing/vehicle traffic, read-only hardware smoke, protocol execution,
any physical clear/reset/write/security/programming, model/runtime changes,
generic framework, new milestone renumbering or Step-15.
## REPORTING REQUIREMENTS
Save `prompts/Step-14-Report.md` entirely inside one plain triple-backtick fence,
no language tag/outside content/nested fences. Include identifier/title,
COMPLETE FOR READINESS or INCOMPLETE/BLOCKED precisely; C/AC source/evidence
mapping, deliverable links, exact proposed subset/no-go/unknowns, local checks
performed, source-ordering self-review/N/A rationale, preservation and issues.
Research can be complete with an explicit unavailable-device prerequisite;
actual physical integration/hardware milestone is still NOT IMPLEMENTED.
Missing authoritative evidence for the proposed subset is a no-go, not assumed
support. Save and STOP for manual review before any implementation assignment.

View File

@ -5,6 +5,11 @@ This supplements the active assignment and existing protected-file/manual
workflow rules. It does not impose new criteria on completed work, authorize workflow rules. It does not impose new criteria on completed work, authorize
agents/automation, or replace repository instructions. agents/automation, or replace repository instructions.
Revision 2026-11-01 applies prospectively from Step-14. The existing "test
pending work" rule missed post-work cleanup/release ordering in Step-13.
The added boundary ledger below makes that review executable, not a claim
that more test cases or longer reports alone prevent defects.
## 1. Before coding: make the contract executable ## 1. Before coding: make the contract executable
Read the assignment, current implementation and its acceptance cases. Establish Read the assignment, current implementation and its acceptance cases. Establish
@ -37,6 +42,15 @@ Do not broaden scope to speculative future features.
hidden defaults, false zero data, swallowed failure or broad exception mask. hidden defaults, false zero data, swallowed failure or broad exception mask.
- When state/lifetime changes, test competing requests while work is pending, - When state/lifetime changes, test competing requests while work is pending,
not only sequential commands after each await completes. not only sequential commands after each await completes.
- Include the completion tail: operation work has returned, but callbacks/
resource release/disposal and the advertised drain task are unfinished.
Ownership remains reserved until cleanup and coherent terminal/task settlement;
reject competing admission with exact unchanged-owner/task/call assertions.
Exercise success, operation failure, cancellation and relevant cleanup failure.
A test holding only the HTTP reply before work returns does not cover this tail.
- Before adding an owner/lifecycle path, inspect accepted local prior art and
use its ordering rather than copying an older incomplete pattern. Keep
callbacks/awaited cleanup outside required state gates and verify lock order.
- Use controlled schedulers/barriers for timing. No arbitrary sleeps/polling as - Use controlled schedulers/barriers for timing. No arbitrary sleeps/polling as
correctness proof. Always release test barriers/resources in finally cleanup. correctness proof. Always release test barriers/resources in finally cleanup.
- Exercise the real implementation under test. Faults may be injected below the - Exercise the real implementation under test. Faults may be injected below the
@ -60,9 +74,28 @@ Do one source-path review of the changed flow:
5. Check every required matrix row has a meaningful assertion and executed result. 5. Check every required matrix row has a meaningful assertion and executed result.
6. Check actual dependencies/Git scope, docs/README, protected files and no secrets. 6. Check actual dependencies/Git scope, docs/README, protected files and no secrets.
For each changed stateful public flow, include this compact ledger in the
required worker report, using actual file/symbol locations and named tests:
| Boundary | Source ordering / invariant | Executed evidence |
|---|---|---|
| Admission | All rejecting checks before dispatch or valid-state/owner mutation | Rejected input/Busy/stale assertions, exact unchanged state and call counts |
| Publication | Identity/source/current owner and complete data verified at commit, not only prechecked | Held publication/supersession and exact output assertions |
| Completion/close | Reserve owner through callback/release/disposal; coherent terminal state and exact completion task; no late successor mutation | Held post-work cleanup/close/reentry assertions, including failure |
| UI/producer-consumer | Real command/render/lifetime uses the same guarded flow, errors observed, canonical values unchanged | Actual rendered/native affected path; mocks identified precisely |
Use N/A with a concrete reason for unaffected boundaries (for example a
research-only assignment), not fabricated tests. Keep this inside the report;
do not create another checklist/notes file or delegate this own review.
Review all touched producer/consumer paths once against the frozen matrix,
not one attractive happy path followed by a blanket "self-review passed".
Fix issues found in this pass inside the current assignment and rerun affected Fix issues found in this pass inside the current assignment and rerun affected
checks. Do not declare COMPLETE after only the first plausible green result or checks. Do not declare COMPLETE after only the first plausible green result or
hand known defects to the orchestrator as recommendations. hand known defects to the orchestrator as recommendations.
The final validation must cover the last relevant production edits; earlier
green runs or an older native binary are not the final changed-path evidence.
Separate original historical evidence from the actual rebuilt affected checks.
This is the worker's own review; do not invoke another agent or automate handoff. This is the worker's own review; do not invoke another agent or automate handoff.
It is a short risk-based check, not a new document/report file or broad unrelated It is a short risk-based check, not a new document/report file or broad unrelated
@ -77,6 +110,13 @@ summary and honest native versus headless coverage.
Use PASSED, FAILED, NOT EXECUTED or BLOCKED precisely. COMPLETE requires all Use PASSED, FAILED, NOT EXECUTED or BLOCKED precisely. COMPLETE requires all
assigned acceptance checks; use INCOMPLETE/BLOCKED for unresolved requirements. assigned acceptance checks; use INCOMPLETE/BLOCKED for unresolved requirements.
Keep historical reports unchanged and report necessary deviations explicitly. Keep historical reports unchanged and report necessary deviations explicitly.
Before returning COMPLETE, reconcile every frozen criterion to a named
executed assertion/output, the boundary ledger where applicable, final command
exits and unresolved issue list. A full-suite count, prose source review or
stub/native substitution cannot close an unsupported row.
Fix tightly coupled findings now, within the same assignment, rather than
shipping known gaps for another lettered step. Do not invent new scope or
silently waive material regressions to reduce the apparent correction count.
After saving the single fenced report, STOP. No future step, orchestration edit, After saving the single fenced report, STOP. No future step, orchestration edit,
automatic agent communication, staging/commit/push or unauthorized system change. automatic agent communication, staging/commit/push or unauthorized system change.

View File

@ -38,6 +38,7 @@ public sealed class CalibrationComparisonSession(CalibrationRehearsalSession reh
private Task? _close; private Task? _close;
private long _generation; private long _generation;
public event Action? Changed; public event Action? Changed;
internal event Action<Task>? ReleasingOwner;
public IReadOnlyList<DiagnosticProviderDescriptor> Providers => diagnostics.Providers; public IReadOnlyList<DiagnosticProviderDescriptor> Providers => diagnostics.Providers;
public CalibrationComparisonSnapshot Snapshot { get { lock (_gate) return _snapshot; } } public CalibrationComparisonSnapshot Snapshot { get { lock (_gate) return _snapshot; } }
@ -306,6 +307,10 @@ public sealed class CalibrationComparisonSession(CalibrationRehearsalSession reh
Exception? error = null; Exception? error = null;
try { await Task.Run(() => work(generation, owner.Token)).ConfigureAwait(false); } try { await Task.Run(() => work(generation, owner.Token)).ConfigureAwait(false); }
catch (Exception exception) { error = exception; } catch (Exception exception) { error = exception; }
try { ReleasingOwner?.Invoke(completion.Task); }
catch (Exception exception) { error = error is null ? exception : new AggregateException(error, exception); }
try { owner.Release(); }
catch (Exception exception) { error = error is null ? exception : new AggregateException(error, exception); }
lock (_gate) lock (_gate)
{ {
_owner = null; _owner = null;
@ -313,9 +318,8 @@ public sealed class CalibrationComparisonSession(CalibrationRehearsalSession reh
_snapshot = _snapshot with { State = error is null ? CalibrationComparisonState.Completed : _snapshot = _snapshot with { State = error is null ? CalibrationComparisonState.Completed :
error is OperationCanceledException ? CalibrationComparisonState.Cancelled : CalibrationComparisonState.Failed, error is OperationCanceledException ? CalibrationComparisonState.Cancelled : CalibrationComparisonState.Failed,
IsBusy = false, Error = error?.Message }; IsBusy = false, Error = error?.Message };
if (error is null) completion.TrySetResult(); else completion.TrySetException(error);
} }
owner.Release();
if (error is null) completion.TrySetResult(); else completion.TrySetException(error);
Changed?.Invoke(); Changed?.Invoke();
} }
} }
@ -360,8 +364,11 @@ public sealed class CalibrationComparisonSession(CalibrationRehearsalSession reh
catch (Exception exception) { error = exception; } catch (Exception exception) { error = exception; }
try { if (drain is not null) await drain.ConfigureAwait(false); } try { if (drain is not null) await drain.ConfigureAwait(false); }
catch (Exception exception) { error = error is null ? exception : new AggregateException(error, exception); } catch (Exception exception) { error = error is null ? exception : new AggregateException(error, exception); }
lock (_gate) _snapshot = _snapshot with { State = CalibrationComparisonState.Closed, IsBusy = false, Error = error?.Message ?? _snapshot.Error }; lock (_gate)
completion.TrySetResult(); {
_snapshot = _snapshot with { State = CalibrationComparisonState.Closed, IsBusy = false, Error = error?.Message ?? _snapshot.Error };
completion.TrySetResult();
}
Changed?.Invoke(); Changed?.Invoke();
} }
public async ValueTask DisposeAsync() { await CloseAsync().ConfigureAwait(false); Changed = null; } public async ValueTask DisposeAsync() { await CloseAsync().ConfigureAwait(false); Changed = null; }

View File

@ -0,0 +1,3 @@
using System.Runtime.CompilerServices;
[assembly: InternalsVisibleTo("AutoForge.IntegrationTests")]

View File

@ -0,0 +1,320 @@
using AutoForge.Application.Features.Vehicles.Calibration;
using AutoForge.Application.Features.Vehicles.Telemetry.Recording;
using AutoForge.Simulation;
namespace AutoForge.IntegrationTests;
public sealed partial class CalibrationRehearsalTests
{
[Theory]
[InlineData(false, "evaluate")]
[InlineData(false, "save")]
[InlineData(true, "evaluate")]
[InlineData(true, "save")]
public async Task ComparisonOwnershipTail_RealCancellationCallbackReservesAdmissionUntilRelease(
bool failReadback, string contender)
{
using var fixture = new Fixture(dynamics: true);
await using var rehearsal = fixture.Session;
var (operation, head) = await PrepareComparisonAsync(fixture);
using var store = new CompletionTailStore(failReadback);
await using var comparison = CreateComparison(fixture, store);
await comparison.EvaluateAsync(operation, fixture.Workspace.Identifier, head.Identifier,
DynamicsDiagnosticTransportFactory.ProviderIdentifier);
var result = comparison.Snapshot.Result;
var releaseEntered = new TaskCompletionSource<Task>(TaskCreationOptions.RunContinuationsAsynchronously);
comparison.ReleasingOwner += completion => releaseEntered.TrySetResult(completion);
store.ReadState = () => comparison.Snapshot;
var save = comparison.SaveCapturesAsync();
Task? cancellation = null;
Task? competing = null;
try
{
await store.ReadbackEntered.Task.WaitAsync(Deadline);
cancellation = Task.Run(() => comparison.CancelAsync());
await store.CallbackEntered.Task.WaitAsync(Deadline);
Assert.True(store.CallbackState!.IsBusy);
store.ReleaseReadback.TrySetResult();
Assert.Same(save, await releaseEntered.Task.WaitAsync(Deadline));
var tail = comparison.Snapshot;
var reads = fixture.Port.ReadCount;
var writes = fixture.Port.WriteCount;
var configurations = fixture.Configurations.List().Count;
CalibrationComparisonException? rejection = null;
try
{
competing = contender == "evaluate"
? comparison.EvaluateAsync(operation, fixture.Workspace.Identifier, head.Identifier,
DynamicsDiagnosticTransportFactory.ProviderIdentifier)
: comparison.SaveCapturesAsync();
}
catch (CalibrationComparisonException exception) { rejection = exception; }
Assert.True(rejection?.Failure == CalibrationComparisonFailure.Busy,
$"tail busy={tail.IsBusy}; state={tail.State}; old drain completed={save.IsCompleted}; " +
$"callback exited={store.CallbackExited}; {contender} admitted={competing is not null}");
Assert.True(tail.IsBusy);
Assert.Equal(CalibrationComparisonState.Saving, tail.State);
Assert.False(save.IsCompleted);
Assert.False(store.CallbackExited);
Assert.Equal(reads, fixture.Port.ReadCount);
Assert.Equal(writes, fixture.Port.WriteCount);
Assert.Equal(configurations, fixture.Configurations.List().Count);
Assert.Equal(2, store.Inner.SaveCalls);
Assert.Same(result, comparison.Snapshot.Result);
}
finally
{
store.ReleaseReadback.TrySetResult();
store.ReleaseCallback.TrySetResult();
await ObserveTailTaskAsync(save);
if (cancellation is not null) await ObserveTailTaskAsync(cancellation);
if (competing is not null) await ObserveTailTaskAsync(competing);
}
Assert.True(store.CallbackExited);
Assert.Equal(1, store.CallbackCalls);
Assert.False(comparison.Snapshot.IsBusy);
Assert.Equal(failReadback ? CalibrationComparisonState.Failed : CalibrationComparisonState.Completed,
comparison.Snapshot.State);
Assert.True(comparison.Snapshot.Saves[0].Persisted && comparison.Snapshot.Saves[0].ReadbackVerified);
Assert.True(comparison.Snapshot.Saves[1].Persisted);
Assert.Equal(!failReadback, comparison.Snapshot.Saves[1].ReadbackVerified);
if (failReadback) Assert.Contains("controlled final readback failure", comparison.Snapshot.Error);
var retry = comparison.SaveCapturesAsync();
Assert.NotSame(save, retry);
await retry.WaitAsync(Deadline);
Assert.Same(retry, comparison.CancelAsync());
Assert.All(comparison.Snapshot.Saves, outcome => Assert.True(outcome.ReadbackVerified));
Assert.Equal(2, store.Inner.SaveCalls);
}
[Theory]
[InlineData(false)]
[InlineData(true)]
public async Task ComparisonOwnershipTail_CloseOrDisposeIncludesOldDrainAfterCompetingSave(bool dispose)
{
using var fixture = new Fixture(dynamics: true);
await using var rehearsal = fixture.Session;
var (operation, head) = await PrepareComparisonAsync(fixture);
using var store = new CompletionTailStore(false);
await using var comparison = CreateComparison(fixture, store);
await comparison.EvaluateAsync(operation, fixture.Workspace.Identifier, head.Identifier,
DynamicsDiagnosticTransportFactory.ProviderIdentifier);
var result = comparison.Snapshot.Result;
var releaseEntered = new TaskCompletionSource<Task>(TaskCreationOptions.RunContinuationsAsynchronously);
var closed = new TaskCompletionSource<(bool DrainSettled, bool CallbackExited)>(
TaskCreationOptions.RunContinuationsAsynchronously);
comparison.ReleasingOwner += completion => releaseEntered.TrySetResult(completion);
store.ReadState = () => comparison.Snapshot;
var save = comparison.SaveCapturesAsync();
comparison.Changed += () =>
{
if (comparison.Snapshot.State == CalibrationComparisonState.Closed)
closed.TrySetResult((save.IsCompleted, store.CallbackExited));
};
Task? cancellation = null;
Task? competing = null;
Task? close = null;
try
{
await store.ReadbackEntered.Task.WaitAsync(Deadline);
cancellation = Task.Run(() => comparison.CancelAsync());
await store.CallbackEntered.Task.WaitAsync(Deadline);
store.ReleaseReadback.TrySetResult();
Assert.Same(save, await releaseEntered.Task.WaitAsync(Deadline));
try { competing = comparison.SaveCapturesAsync(); }
catch (CalibrationComparisonException exception) { Assert.Equal(CalibrationComparisonFailure.Busy, exception.Failure); }
if (competing is not null)
{
await competing.WaitAsync(Deadline);
close = dispose ? comparison.DisposeAsync().AsTask() : comparison.CloseAsync();
Assert.False(close.IsCompleted,
$"replacement admitted=True; old drain completed={save.IsCompleted}; " +
$"callback exited={store.CallbackExited}; Close/Dispose completed={close.IsCompleted}; state={comparison.Snapshot.State}");
}
else
{
var closeRequested = new TaskCompletionSource(TaskCreationOptions.RunContinuationsAsynchronously);
close = Task.Run(async () =>
{
closeRequested.TrySetResult();
if (dispose) await comparison.DisposeAsync(); else await comparison.CloseAsync();
});
await closeRequested.Task.WaitAsync(Deadline);
Assert.False(save.IsCompleted);
Assert.False(closed.Task.IsCompleted);
}
}
finally
{
store.ReleaseReadback.TrySetResult();
store.ReleaseCallback.TrySetResult();
await ObserveTailTaskAsync(save);
if (cancellation is not null) await ObserveTailTaskAsync(cancellation);
if (competing is not null) await ObserveTailTaskAsync(competing);
if (close is not null) await close.WaitAsync(Deadline);
}
var terminal = await closed.Task.WaitAsync(Deadline);
Assert.True(terminal.DrainSettled);
Assert.True(terminal.CallbackExited);
Assert.Equal(1, store.CallbackCalls);
Assert.Equal(CalibrationComparisonState.Closed, comparison.Snapshot.State);
Assert.False(comparison.Snapshot.IsBusy);
Assert.Same(result, comparison.Snapshot.Result);
Assert.False(result!.IsCurrent);
Assert.Same(comparison.CloseAsync(), comparison.CloseAsync());
Assert.Throws<CalibrationComparisonException>(() => { _ = comparison.SaveCapturesAsync(); });
await using var fresh = CreateComparison(fixture, new ComparisonMemoryStore());
await fresh.EvaluateAsync(operation, fixture.Workspace.Identifier, head.Identifier,
DynamicsDiagnosticTransportFactory.ProviderIdentifier);
Assert.Equal(CalibrationComparisonState.Completed, fresh.Snapshot.State);
Assert.NotSame(result, fresh.Snapshot.Result);
Assert.Equal(1, fixture.Port.WriteCount);
}
[Fact]
public async Task ComparisonOwnershipTail_CancelledEvaluationKeepsExactDrainUntilCleanupAndThenRetries()
{
var factory = new ComparisonSignalFaultFactory("held");
using var fixture = new Fixture(dynamics: true, dynamicsFactory: factory);
await using var rehearsal = fixture.Session;
var (operation, head) = await PrepareComparisonAsync(fixture);
await using var comparison = CreateComparison(fixture, new ComparisonMemoryStore());
var releaseEntered = new TaskCompletionSource<Task>(TaskCreationOptions.RunContinuationsAsynchronously);
var releaseCleanup = new TaskCompletionSource(TaskCreationOptions.RunContinuationsAsynchronously);
var releaseCalls = 0;
void HoldRelease(Task completion)
{
Interlocked.Increment(ref releaseCalls);
releaseEntered.TrySetResult(completion);
releaseCleanup.Task.GetAwaiter().GetResult();
}
comparison.ReleasingOwner += HoldRelease;
var evaluation = comparison.EvaluateAsync(operation, fixture.Workspace.Identifier, head.Identifier,
DynamicsDiagnosticTransportFactory.ProviderIdentifier);
Task? cancellation = null;
try
{
await factory.Entered.Task.WaitAsync(Deadline);
cancellation = comparison.CancelAsync();
Assert.Same(evaluation, cancellation);
factory.Release.TrySetResult();
Assert.Same(evaluation, await releaseEntered.Task.WaitAsync(Deadline));
var tail = comparison.Snapshot;
Assert.True(tail.IsBusy,
$"cancelled evaluation tail busy={tail.IsBusy}; state={tail.State}; old drain completed={evaluation.IsCompleted}");
Assert.Equal(CalibrationComparisonState.Evaluating, tail.State);
Assert.False(evaluation.IsCompleted);
Assert.Same(evaluation, comparison.CancelAsync());
Assert.Throws<CalibrationComparisonException>(() => { _ = comparison.SaveCapturesAsync(); });
Assert.Null(tail.Result);
}
finally
{
factory.Release.TrySetResult();
releaseCleanup.TrySetResult();
await ObserveTailTaskAsync(evaluation);
if (cancellation is not null) await ObserveTailTaskAsync(cancellation);
comparison.ReleasingOwner -= HoldRelease;
}
await Assert.ThrowsAnyAsync<OperationCanceledException>(() => evaluation);
Assert.Equal(1, releaseCalls);
Assert.Equal(1, factory.Disposals);
Assert.Equal(CalibrationComparisonState.Cancelled, comparison.Snapshot.State);
Assert.False(comparison.Snapshot.IsBusy);
Assert.False(comparison.Snapshot.PartialRun!.IsComplete);
var next = comparison.EvaluateAsync(operation, fixture.Workspace.Identifier, head.Identifier,
DynamicsDiagnosticTransportFactory.ProviderIdentifier);
Assert.NotSame(evaluation, next);
await next.WaitAsync(Deadline);
Assert.Equal(CalibrationComparisonState.Completed, comparison.Snapshot.State);
Assert.True(comparison.Snapshot.Result!.IsCurrent);
}
[Theory]
[InlineData(false)]
[InlineData(true)]
public async Task ComparisonOwnerCleanup_FailureSettlesWaitersAndPreservesPrimaryAndSaveEvidence(bool primaryFailure)
{
using var fixture = new Fixture(dynamics: true);
await using var rehearsal = fixture.Session;
var (operation, head) = await PrepareComparisonAsync(fixture);
var store = new ComparisonMemoryStore { FailSaveCall = primaryFailure ? 2 : null };
await using var comparison = CreateComparison(fixture, store);
await comparison.EvaluateAsync(operation, fixture.Workspace.Identifier, head.Identifier,
DynamicsDiagnosticTransportFactory.ProviderIdentifier);
var result = comparison.Snapshot.Result;
var failure = new IOException("controlled completion cleanup failure");
void FailRelease(Task _) => throw failure;
comparison.ReleasingOwner += FailRelease;
var save = comparison.SaveCapturesAsync();
try
{
var error = await Assert.ThrowsAnyAsync<Exception>(() => save.WaitAsync(Deadline));
Assert.IsNotType<TimeoutException>(error);
if (primaryFailure)
{
var aggregate = Assert.IsType<AggregateException>(error);
Assert.Contains(aggregate.InnerExceptions, item => item is CalibrationComparisonException);
Assert.Contains(failure, aggregate.InnerExceptions);
}
else Assert.Same(failure, error);
Assert.Equal(CalibrationComparisonState.Failed, comparison.Snapshot.State);
Assert.Contains(failure.Message, comparison.Snapshot.Error);
Assert.False(comparison.Snapshot.IsBusy);
Assert.Same(result, comparison.Snapshot.Result);
Assert.True(comparison.Snapshot.Saves[0].ReadbackVerified);
Assert.Equal(!primaryFailure, comparison.Snapshot.Saves[1].Persisted);
}
finally { comparison.ReleasingOwner -= FailRelease; }
store.FailSaveCall = null;
await comparison.SaveCapturesAsync().WaitAsync(Deadline);
Assert.All(comparison.Snapshot.Saves, outcome => Assert.True(outcome.ReadbackVerified));
}
private static async Task ObserveTailTaskAsync(Task task)
{
try { await task.WaitAsync(Deadline); }
catch (Exception exception) when (exception is CalibrationComparisonException or OperationCanceledException) { }
}
private sealed class CompletionTailStore(bool failReadback) : ITelemetryArchiveStore, IDisposable
{
public ComparisonMemoryStore Inner { get; } = new();
public Func<CalibrationComparisonSnapshot>? ReadState { get; set; }
public CalibrationComparisonSnapshot? CallbackState { get; private set; }
public TaskCompletionSource ReadbackEntered { get; } = new(TaskCreationOptions.RunContinuationsAsynchronously);
public TaskCompletionSource ReleaseReadback { get; } = new(TaskCreationOptions.RunContinuationsAsynchronously);
public TaskCompletionSource CallbackEntered { get; } = new(TaskCreationOptions.RunContinuationsAsynchronously);
public TaskCompletionSource ReleaseCallback { get; } = new(TaskCreationOptions.RunContinuationsAsynchronously);
public bool CallbackExited { get; private set; }
public int CallbackCalls { get; private set; }
private CancellationTokenRegistration _registration;
private int _reads;
public Task SaveAsync(TelemetryCaptureSnapshot capture, CancellationToken cancellationToken = default)
{
if (Inner.SaveCalls == 0)
_registration = cancellationToken.Register(() =>
{
CallbackCalls++;
CallbackState = ReadState!();
CallbackEntered.TrySetResult();
ReleaseCallback.Task.GetAwaiter().GetResult();
CallbackExited = true;
});
return Inner.SaveAsync(capture, cancellationToken);
}
public async Task<TelemetryCaptureSnapshot> LoadAsync(string captureId, CancellationToken cancellationToken = default)
{
var readback = await Inner.LoadAsync(captureId);
if (++_reads != 2) return readback;
ReadbackEntered.TrySetResult();
await ReleaseReadback.Task;
if (failReadback)
throw new TelemetryArchiveException(TelemetryArchiveFailure.ReadFailure, "controlled final readback failure");
return readback;
}
public Task<IReadOnlyList<TelemetrySavedCapture>> ListAsync(CancellationToken cancellationToken = default) =>
Inner.ListAsync(cancellationToken);
public void Dispose() => _registration.Dispose();
}
}