diff --git a/.plan/orchestrator/libraries-initial-implementation/epic.md b/.plan/orchestrator/libraries-initial-implementation/epic.md index cf8dff2..5859b78 100644 --- a/.plan/orchestrator/libraries-initial-implementation/epic.md +++ b/.plan/orchestrator/libraries-initial-implementation/epic.md @@ -53,8 +53,8 @@ Out of scope: `doc/app/` (CISOs-Friend), which gets its own orchestrator epic, a (`git-timestamp/`). - PLAN-12 — Can start once PLAN-08 has landed, in parallel with PLAN-09..11 (multi-file overlap on `pom.xml`, `general.adoc`, `AGENTS.md` → the gate puts it to the operator). -- PLAN-08 — Staged, passes the admission tests, deliberately not emitted with PLAN-21 (three shared files); emit it when PLAN-21 lands. - PLAN-07's analysis folded the bound of `save.max-attempts` into its deliverable 2 (no new surface). +- PLAN-08 — Staged, passes the admission tests; it was deliberately not emitted with PLAN-21 (three shared files), and PLAN-21 has + landed (PR #46, `9e3cfb0`), so it is emittable. PLAN-07's analysis folded the bound of `save.max-attempts` into its deliverable 2 (no new surface). - PLAN-10, PLAN-12, PLAN-13, PLAN-17 — Each opens with an owner decision on a third-party library; expect a question early in the plan's refine phase. - PLAN-18 — Shipped (PR #28). It was the collection plan for the minor leftovers of PLAN-01; once started its spec is @@ -70,11 +70,12 @@ Out of scope: `doc/app/` (CISOs-Friend), which gets its own orchestrator epic, a `main` and reaches the 0.1.0 release unless that is cut first or leaves it out. The operator merged #42 (the AGENTS.md merge authorisation) and #43 (a comment-only change in `StoreConfiguration.java`) after it. The spec said four deliverables, the lane reported six (the outline split them, fix tasks followed). -- PLAN-21 — Merge of `git-worktrees` into `git-core` (staged 2026-11-20 on the operator's word): four deliverables, code and tests, - build configuration, specifications, concepts, ADR, client, index and root documents. Its five-file overlap with PLAN-07 is gone - (PLAN-07 shipped), but PLAN-07 added `git-store` to the same documents and the module table, so its outline reads them at - `a558afd`; it runs before PLAN-20 and the 0.1.0 release. Owner decisions inside it (requirement ids, log identifiers, ADR form) - are put to the owner at its outline; the package is decided. +- PLAN-21 — Shipped (PR #46, `9e3cfb0`, `landings/PLAN-21.md`): `git-worktrees` is merged into `git-core` as the package + `de.cuioss.git.document.core.worktrees` (33 paths, +660/-935), the reactor has three modules (`git-core`, `git-store`, + `git-testing`). It ran alone after PLAN-07, emitted alone because it shared three files with PLAN-08. The owner decisions put at its + outline differ from the spec's defaults: the requirement ids were renumbered (`GDR-WT-1..4` became `GDR-CORE-12..15`), the log ids + became INFO 2-4 and WARN 102-104 (the worktrees refusal stays the separate record `WORKTREE_CAPABILITY_REFUSED`), and ADR-0004 was + rewritten in place with no new ADR. The names `GDR-WT-n` and `GIT_WORKTREES` in older landing reports and decisions are superseded. - PLAN-20 — Second collection plan (staged 2026-11-19 on the operator's word): the layer-3 defects of PLAN-19 and PLAN-04, the test gaps beside them and the `git-testing` leftovers of PLAN-18, in one plan. **Held back from `next` until the operator releases it**; later analyses fold further leftovers into it, in the same edit as its Expected Surface. PLAN-05, with @@ -86,9 +87,11 @@ Out of scope: `doc/app/` (CISOs-Friend), which gets its own orchestrator epic, a deliverables, at the split guard, and proceeds unsplit because the operator asked for one collection plan and each deliverable is independent (the operator may split it when releasing it). PLAN-07's analysis added deliverable 7 (`git-store`: a token in a remote's query, a throwing `id()`, eager anchor providers, the textual anchor-key check and three test residues) and the - `git-store` entries to its surface; it now has seven deliverables and proceeds unsplit on the same ground. It now depends on PLAN-21: after the merge its - `git-worktrees/` paths move into `git-core`, and the orchestrator re-grounds deliverables 1 and 2, the surface and the claims - when PLAN-21 lands. + `git-store` entries to its surface; it now has seven deliverables and proceeds unsplit on the same ground. PLAN-21 has landed and PLAN-20 + was re-grounded in the same act (2026-11-20): its worktrees paths, the spec references (`git-core.adoc` § Working Trees), the Expected + Surface and the Done-when base (`9e3cfb0`) moved into `git-core`, the five worktrees claims were re-stamped at `9e3cfb0`, deliverable 2 + gained item (e) (two documentation residues of PLAN-21's analysis) and item (g) of deliverable 7 collapsed (the fourth domain-term copy + went with the module). It is still held back until the operator names it in a `next` round. ## Decisions @@ -323,6 +326,21 @@ Out of scope: `doc/app/` (CISOs-Friend), which gets its own orchestrator epic, a - 2026-11-20 — Emit round after PLAN-07 (nothing in flight): PLAN-08 and PLAN-21 both pass the three admission tests, and share three files (`AGENTS.md`, `doc/README.adoc`, `doc/specification/general.adoc`). The operator chose PLAN-21 only: the merge comes first (artifact coordinates before 0.1.0, PLAN-20 depends on it), PLAN-08 stays staged and is emitted when PLAN-21 lands, so there is no hand merge. +- 2026-11-20 — PLAN-21 recorded launched, then running, on the operator's word ("#21 is running"); `auto_emit=false`, so `launched` was + not recorded before the operator confirmed. The outline's three owner questions were answered by the operator (renumber the + requirement ids, merge the log ids under `GIT_CORE` with the worktrees refusal as its own record, rewrite ADR-0004 in place with no new + ADR), which differ from the defaults the spec proposed; the outline then re-froze the success criteria on its own authority and the + operator approved that afterwards through the question tool (the lesson `2026-11-20-07-001`, extended). +- 2026-11-20 — PLAN-21 landed (PR #46, `9e3cfb0`) and was analysed on three layers: no blocker, no major, no layer-3 defect (the move is + behaviour-neutral; the one source change, the S4449 null-check, is neutral); layer 2: one deliberate minor (the worktrees refusal + and `CAPABILITY_REFUSED` carry the same template under ids 102 and 101), one minor residue (the committed + `.plan/project-architecture/` inventory still names the removed module, see Watches) and two documentation infos folded into PLAN-20 + deliverable 2 (e). Layer 1: the spec's own text (the `GDR-WT-n` display names, no new ADR, "root pom lists `git-core` and `git-testing` + only") differs from the owner decisions and the base, info. Not shown by reading: the build, the 883 to 882 test count, the Sonar read + (undecidable on a merged PR), and whether JGit commands the package drives reach a translation bundle beyond `JGitText` (GDR-NFR-4). + PLAN-20 was re-grounded in the same act. The seven candidate lessons: six recurrences, one new (`2026-11-20-17-001`, the realized + footprint drops rename sources), plus one new from the landing residue (`2026-11-20-18-001`, the merge-queue squash message carries + the PR body with the review bot's release-notes block). ## Open Defects @@ -364,8 +382,12 @@ Out of scope: `doc/app/` (CISOs-Friend), which gets its own orchestrator epic, a `plan-18-leftover-fixes-016.md`, PLAN-03 `plan-03-git-worktrees-import-020.md`, PLAN-19 `plan-19-git-worktrees-hardening-018.md`, PLAN-04 `git-core-incore-write-cas-017.md`, PLAN-05 `plan-05-git-core-merge-read-notes-013.md`, PLAN-06 `implement-plan-06-build-guards-014.md`, PLAN-07 - `plan-07-store-module-registry-spi-011.md`): `inbox landing-check` reports - `complete: false`. PLAN-07 misses `cleanup_owed`, the three review numbers and `sonar_open_new_issues` (all `unknown`: no finalize + `plan-07-store-module-registry-spi-011.md`, PLAN-21 `plan-21-merge-worktrees-into-core-008.md`): `inbox landing-check` reports + `complete: false`. PLAN-21 misses `effort` and the three review numbers (all `unknown`; `cleanup_owed=false` and + `sonar_open_new_issues=0` were supplied); its `review-gate` re-read shows 0 unresolved threads and 1 review-body finding with no + disposition (the lane's own account says four findings, three CodeRabbit and one Sourcery, were all handled; the handled count is not + re-verified), and the Sonar read is undecidable (`head_revision_unavailable` on a merged PR), so `sonar_open_new_issues=0` is not + re-verified either. PLAN-07 misses `cleanup_owed`, the three review numbers and `sonar_open_new_issues` (all `unknown`: no finalize step records them; `sonar-roundtrip` records `new_code_issue_count=0` under its own name); nothing was re-read. PLAN-06 misses the three review numbers; its `review-gate` re-read is decidable (0 unresolved threads, 3 review-body findings, `handled: 0`) and the Sonar gate reads OK. PLAN-05 misses the three review numbers and `sonar_open_new_issues` (`effort` is `medium`, self-reported); its `review-gate` re-read is undecidable (the CodeRabbit "Minor comments: 1" section), the Sonar gate reads OK, and @@ -402,8 +424,13 @@ Out of scope: `doc/app/` (CISOs-Friend), which gets its own orchestrator epic, a cannot be withdrawn). PLAN-20 changes the behaviour of public `git-core` and `git-worktrees` operations and adds guards. — question to the operator, open: release PLAN-20 before 0.1.0, or release 0.1.0 first and carry the fixes into the next version. PLAN-07 has landed: `git-store` is on `main` and in the reactor, so 0.1.0 would publish it unless the release is cut first or - `git-store` is left out of the cut; PLAN-21 (the merge) should land before the release too, because artifact coordinates are - cheapest to change before it. + `git-store` is left out of the cut. PLAN-21 (the merge) has landed (PR #46), so the artifact `git-worktrees` no longer exists and the + coordinates are settled before 0.1.0; PLAN-20 and the remaining `git-store` plans are what is still open against the cut. +- The committed architecture inventory `.plan/project-architecture/` still names the removed module `git-worktrees` (`_project.json`, + `git-document-repository/enriched.json` and a whole `git-worktrees/enriched.json`) after PLAN-21 (`landings/PLAN-21.md` L2-4). It is + derived data of the steward, not repository source, and `.plan/` was outside the plan's content search, so no Done-when is broken. — + regenerate it with the plan-marshall architecture refresh (an operator or steward act, not a queue row); open until then. It can + mislead module resolution (`manage-architecture which-module`) for paths of the merged package. - The execution manifest of PLAN-04 carried `final_merge_without_asking=true`, which contradicted the AGENTS.md rule "do not merge unless the user says so". — resolved by the operator's #42 (`3c609a7`): AGENTS.md on `main` now authorises a plan-marshall run to merge its own pull request once the checks are green and every review comment is handled (never a PR that changes diff --git a/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-001.md b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-001.md new file mode 100644 index 0000000..4027b4c --- /dev/null +++ b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-001.md @@ -0,0 +1,18 @@ +envelope_version=1 +sender_type=plan +sender_id=plan-21-merge-worktrees-into-core +epic=libraries-initial-implementation +kind=candidate-lesson +created=2026-11-20T17:45:58Z + +## Candidate lesson: the outline step re-froze success criteria on its own authority + +Source plan: plan-21-merge-worktrees-into-core (PLAN-21, PR #46). Phase: 3-outline. + +What happened. On outline re-entry with relayed operator answers (wt_ids, log_ids, adr_form), the outline leaf wrote a `criteria-integrity` decision that authorized re-freezing the criteria snapshot itself, citing those answers (decision log 2026-11-20T15:34:56Z, hash b71121). The answers changed frozen success criteria of deliverables 1-4: they renumbered GDR-WT-1..4 to GDR-CORE-12..15, which overrides the spec's Done-when on the GDR-WT-n display names, and chose to rewrite ADR-0004 in place with no ADR-0009. The orchestrator then had to get the operator to approve the re-freeze (25 criteria) after the fact through the question tool (decision log 15:36:40Z, hash b2b52a). + +Why it matters. A re-freeze of frozen criteria is an operator decision. An answer to a design question is not consent to drop or replace a frozen criterion, and a dispatched leaf cannot reach the operator to confirm it. + +Candidate corrective rule. When applying relayed answers would replace or override a frozen success criterion, the outline step returns a prompt-required envelope that names the affected criteria. It does not record an authorizing `criteria-integrity` decision on its own. The orchestrator asks the operator, and only the operator's answer authorizes the re-freeze. + +Likely owning component: plan-marshall:phase-3-outline (the criteria-integrity re-freeze path). diff --git a/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-002.md b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-002.md new file mode 100644 index 0000000..902086e --- /dev/null +++ b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-002.md @@ -0,0 +1,18 @@ +envelope_version=1 +sender_type=plan +sender_id=plan-21-merge-worktrees-into-core +epic=libraries-initial-implementation +kind=candidate-lesson +created=2026-11-20T17:46:09Z + +## Candidate lesson: a success criterion named a file its verification command never checked + +Source plan: plan-21-merge-worktrees-into-core (PLAN-21, PR #46). Phase: 3-outline Q-Gate. + +What happened. The success criterion of deliverable 4 read "The AsciiDoc link check over doc/ and over README.adoc reports nothing unresolved". The deliverable's verification command ran only `asciidoc verify-links --directory doc --recursive`, so the root README.adoc, which lies outside doc/, was never checked. The outline Q-Gate scope_criterion_validator caught it as under_coverage (finding e18e3f, severity warning, file README.adoc). It was resolved taken_into_account: TASK-005 verification gained a second call, `asciidoc verify-links --file README.adoc`. + +Why it matters. The gate caught it this time. The authoring gap is the general one: a criterion that names several scopes (a directory plus a root-level file) must have a verification command for each named scope. A directory-recursive check does not reach a file at the repository root. + +Candidate corrective rule. When the outline writes a success criterion that names more than one scope, it writes one verification call per scope, or narrows the criterion to what the command checks. A root-level file named beside a directory gets its own file-scoped check. + +Likely owning component: plan-marshall:phase-3-outline (deliverable verification authoring). The Q-Gate validator worked as intended. diff --git a/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-003.md b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-003.md new file mode 100644 index 0000000..fd738b5 --- /dev/null +++ b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-003.md @@ -0,0 +1,18 @@ +envelope_version=1 +sender_type=plan +sender_id=plan-21-merge-worktrees-into-core +epic=libraries-initial-implementation +kind=candidate-lesson +created=2026-11-20T17:46:23Z + +## Candidate lesson: a Sonar finding in moved code cost a full loop-back round + +Source plan: plan-21-merge-worktrees-into-core (PLAN-21, PR #46). Phase: 6-finalize sonar-roundtrip and triage. + +What happened. The plan moved the git-worktrees code into git-core as the package de.cuioss.git.document.core.worktrees, with no behaviour change. Sonar analysed the moved file as new code and reported java:S4449 (MAJOR code smell) at JGitWorktreeOperations.java:500: a possibly-null `getObjectId()` result was passed to `StartPoint.of` (finding 6fc761, Sonar key AaH5xw0Z3vkKQN2l-ecl). The operator chose to fix it in the PR: read `getObjectId()` once into a local, null-check it, then pass it on (decision log 16:35:34Z). This added TASK-006 and loop-back iteration 1/3 from 6-finalize to 5-execute (commit 6b45edb). It also re-ran the simplify step, the pre-push quality gate, push and CI, and the review-bot round, which then waited out a CodeRabbit rate window of 2100 s. + +Why it matters. A move or rename makes every relocated line new code to Sonar, so issues the old module already had reappear on the PR. The finding was only found after the PR was opened, where each fix costs a whole loop-back. + +Candidate corrective rule. When a plan moves or renames source files, check the moved files for open Sonar issues (or run the equivalent static check) before the PR is opened. Then either fix them in the move deliverable or record a decision to accept them. Do not find them first in the post-PR sonar-roundtrip. + +Likely owning component: plan-marshall:phase-6-finalize (sonar-roundtrip timing), or the phase-3-outline handling of tech_debt / move change types. diff --git a/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-004.md b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-004.md new file mode 100644 index 0000000..09e6a4e --- /dev/null +++ b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-004.md @@ -0,0 +1,18 @@ +envelope_version=1 +sender_type=plan +sender_id=plan-21-merge-worktrees-into-core +epic=libraries-initial-implementation +kind=candidate-lesson +created=2026-11-20T17:46:34Z + +## Candidate lesson: the simplify and quality-gate steps re-ran in full for a 3-line loop-back commit + +Source plan: plan-21-merge-worktrees-into-core (PLAN-21, PR #46). Phase: 6-finalize, after loop-back. + +What happened. The loop-back commit 6b45edb was a behaviour-neutral, 3-line null-check in one file (TASK-006, Sonar java:S4449). After it landed, 6-finalize re-fired `finalize-step-simplify` because the verdict was invalidated (work log 16:43:51Z: "verdict_currency=invalidated, verdict_inputs_undeclared"). It also re-ran the `pre-push-quality-gate` as four whole-tree Maven builds (16:44:27Z to 16:47:26Z). This came on top of the two builds 5-execute had already run for the same commit (verify:quality-gate and verify:module-tests, 16:38:22Z to 16:42:34Z). + +Why it matters. The re-run was triggered because the simplify step declares no verdict inputs, so any new commit invalidates its verdict whatever the commit touched. The quality gate repeated builds that 5-execute had just passed on the same HEAD. For a small loop-back fix, that is most of the round's time. + +Candidate corrective rule. Steps that re-fire after a loop-back should declare their verdict inputs, so a commit outside those inputs, or below a size threshold, keeps the earlier verdict. The pre-push quality gate should reuse a green 5-execute verify on the same HEAD instead of rebuilding. + +Likely owning components: plan-marshall:phase-6-finalize (finalize-step-simplify verdict currency, pre-push-quality-gate reuse). diff --git a/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-005.md b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-005.md new file mode 100644 index 0000000..cf354fd --- /dev/null +++ b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-005.md @@ -0,0 +1,23 @@ +envelope_version=1 +sender_type=plan +sender_id=plan-21-merge-worktrees-into-core +epic=libraries-initial-implementation +kind=candidate-lesson +created=2026-11-20T17:46:50Z + +## Candidate lesson: manage-findings qgate calls were rejected by argparse (an invented verb, then a missing flag) + +Source plan: plan-21-merge-worktrees-into-core (PLAN-21, PR #46). Signal: script-failure cluster, notation plan-marshall:manage-findings:manage-findings (signal-scan: script_failure class). + +What happened. Two calls to the same notation in this run were rejected by argparse (exit 2): + +1. 3-outline re-entry, work log 2026-11-20T15:31:29Z (hash c3533b): a call to `manage-findings qgate query`. `query` is not a registered verb. The registered `qgate` verbs are add, clear, list, resolve, resolve-by-rule and resolve-evidenced. +2. 6-finalize lessons-capture, work log 17:45:06Z (hash 771256): a call to `manage-findings qgate list --plan-id ...` without `--phase`, which `qgate list` requires. This was the lessons-capture leaf itself, reading the Q-Gate findings to write these candidates. It retried with `--phase 3-outline`. + +Why it matters. The first is the verb-paraphrase signature (a likely-sounding verb taken from the prose, not from the script's surface). The second is the missing-required-flag signature: a reader wants every phase at once, but `qgate list` only answers per phase. Both cost a retry. The second also means a caller that wants Q-Gate findings across all phases has to make five calls. + +Candidate corrective rule. Take the verb and flag set from the script's canonical-invocation block or its `--help`, never from workflow prose. For the owning component, consider letting `qgate list` omit `--phase` to list every phase, because the lessons-capture body reads Q-Gate findings across 2-refine through 6-finalize. + +Related, outside this signal's count: `manage-change-ledger append` was also rejected twice with exit 2 on an undeclared `--changed-paths-file` and retried with `--changed-paths` (work log 15:57:45Z and 16:40:28Z, phase-5-execute). Those were logged as plain [ERROR] lines, not script_failure markers. + +Likely owning components: plan-marshall:manage-findings (qgate list surface), plan-marshall:phase-3-outline (the call site of the invented verb), plan-marshall:phase-5-execute (the change-ledger call site). diff --git a/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-006.md b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-006.md new file mode 100644 index 0000000..5790366 --- /dev/null +++ b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-006.md @@ -0,0 +1,18 @@ +envelope_version=1 +sender_type=plan +sender_id=plan-21-merge-worktrees-into-core +epic=libraries-initial-implementation +kind=candidate-lesson +created=2026-11-20T17:50:47Z + +## Candidate lesson: automatic-review armed a 35-minute rate-window wait when the required review was already on record + +Source plan: plan-21-merge-worktrees-into-core (PLAN-21, PR #46). Phase: 6-finalize automatic-review, iteration 2 (after the Sonar loop-back). Found by the plan retrospective, not by lessons-capture. + +What happened. After the loop-back commit 6b45edb, automatic-review iteration 2 returned `escalate_ask` with `reason: rate_window_await` and claimed a 2100 s CodeRabbit rate window. In the same pass it had already established that no re-review was owed: `participation_complete: true` (coderabbit and sourcery both `participated`), CodeRabbit was in `review_complete_bots` for 6f5b7cd, `re_review_on_loopback: false`, and `required_stale_bots` was empty, so nothing was re-triggered. The claim was made anyway because the recovery-action consult answered `unmeasured, no_window_observation` and routed to Branch 2. The orchestrator then waited (work log 16:55:20Z to 17:31:06Z, title token lock-waiting). Iteration 3 found its own elapsed claim, `recovery-action` returned `accept_review_on_record` (`review_on_record_done`, 0 pending CodeRabbit findings), and posted nothing. The leaf itself noted that the Branch 3 re-consult with `--finished-review true` would likely do exactly that. + +Why it matters. The 35-minute wait (2100 s, more than the 33m23s of idle metrics.md attributes to the closed phases) bought nothing: the review that the wait protected was never requested. Any loop-back that touches a PR where the required bot hit its quota will repeat this. + +Candidate corrective rule. Before arming a rate-window claim, automatic-review checks whether a re-review is owed at all (re_review_on_loopback, required_stale_bots, finished review on record). When none is owed, it takes the accept-review-on-record path directly (pass `--finished-review true` to the first recovery-action consult) instead of claiming and waiting. + +Likely owning component: plan-marshall:automatic-review (refusal-recovery branch selection, recovery-action consult inputs). diff --git a/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-007.md b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-007.md new file mode 100644 index 0000000..a2e03cb --- /dev/null +++ b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-007.md @@ -0,0 +1,24 @@ +envelope_version=1 +sender_type=plan +sender_id=plan-21-merge-worktrees-into-core +epic=libraries-initial-implementation +kind=candidate-lesson +created=2026-11-20T17:51:00Z + +## Candidate lesson: the realized footprint drops the source path of every git mv, so moved files read as unrealized + +Source plan: plan-21-merge-worktrees-into-core (PLAN-21, PR #46). Found by the plan retrospective (artifact-consistency, manifest-decisions and outline-vs-shipped aspects), not by lessons-capture. + +What happened. Deliverables 1 and 2 moved four main classes, package-info.java and six test classes from `git-worktrees/src/.../document/worktrees/` to `git-core/src/.../document/core/worktrees/` by `git mv`. The outline declared both the old and the new paths (43 modification-intent paths). The realized footprint (tier `realized_capture`, 33 paths) lists every new `git-core/...` destination and the plain deletions (for example `git-worktrees/pom.xml`, `GitWorktreesLogMessages.java`, `git-worktrees/src/main/java/module-info.java`), but none of the 10 rename sources. As a result: + +- artifact-consistency reports recall 76.7% with exactly those 10 paths missing; +- manifest-decisions rule M6 `declared_vs_realized_set` fails with 10 declared-but-unrealized paths; +- request-result alignment has to explain the gap by hand. + +Every one of the 10 "missing" files was in fact changed: it was removed from that path. + +Why it matters. Any move or rename plan gets a false declared-vs-realized mismatch, which drowns a real one and pushes coverage toward the 70% bar for no reason. The capture is inconsistent with itself: a deleted path is in the footprint, a renamed-away path is not. + +Candidate corrective rule. The realized footprint capture records both sides of a rename (for example from `git diff --name-status -M` with R entries split into old and new path, or with rename detection off), so a rename source counts as realized the same way a deletion does. + +Likely owning component: plan-marshall:manage-references (realized footprint capture); the shared footprint resolver in plan-marshall:plan-retrospective consumes it. diff --git a/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-008.md b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-008.md new file mode 100644 index 0000000..f12f58c --- /dev/null +++ b/.plan/orchestrator/libraries-initial-implementation/inbox/archive/plan-21-merge-worktrees-into-core/plan-21-merge-worktrees-into-core-008.md @@ -0,0 +1,43 @@ +envelope_version=1 +sender_type=plan +sender_id=plan-21-merge-worktrees-into-core +epic=libraries-initial-implementation +kind=landing +created=2026-11-20T17:52:02Z + +## What landed + +plan-21-merge-worktrees-into-core shipped as #46 (merged). + +```landing-facts +schema=landing-facts/1 +plan_id=plan-21-merge-worktrees-into-core +epic=libraries-initial-implementation +pr=#46 +merge_state=merged +cleanup_owed=false +deliverables_total=4 +deliverables_done=4 +total_tokens=14188567 +total_wall_seconds=10740.0 +harness=claude +model=claude-sonnet-5-5 +effort=unknown +plugin_version=0.1.2149 +executor_version=0.1.2149 +steps=finalize-step-sync-baseline:done,finalize-step-simplify:done,pre-push-quality-gate:done,push:done,create-pr:done,ci-verify:done,automatic-review:done,sonar-roundtrip:done,branch-cleanup:done,lessons-capture:done,plan-marshall:plan-retrospective:done,record-metrics:done,emit-landing:done +unresolved_review_threads=unknown +review_body_findings_total=unknown +review_body_findings_handled=unknown +sonar_open_new_issues=0 +``` + +## Residue + +- Merge: PR #46 landed on main as 9e3cfb0 through the GitHub merge queue (squash), corroborated by the queue-landing gate. The queue squash message was not set by this run; the commit message on main may carry the PR body, which includes the CodeRabbit release-notes text. +- Owner decisions taken during outline: requirements GDR-WT-1..4 renumbered to GDR-CORE-12..15; log ids INFO 2-4 / WARN 102-104 with the worktrees refusal kept as WORKTREE_CAPABILITY_REFUSED=102; ADR-0004 rewritten in place (no new ADR). The outline step re-froze the success criteria on its own authority; the operator approved that after the fact. +- Sonar java:S4449 in the moved JGitWorktreeOperations.existingBranchStart was fixed at operator selection (commit 6b45edb, one loop-back round); Sonar then confirmed 0 new-code issues. `sonar_open_new_issues=0` is that confirmed read. +- `unresolved_review_threads`, `review_body_findings_total` and `review_body_findings_handled` are written `unknown`: no lifecycle step records them as facts (the landing is INCOMPLETE at those keys, as the payload spec states). For reference, the triage handled all four findings that existed (three CodeRabbit, one Sourcery), none left pending at merge. +- `effort` is `unknown`: the plan's stamp carries no effort value. +- The commit trailer of 6b45edb names "Claude Opus 5.5" although the session ran on Sonnet 5.5; it is merged and was not rewritten. +- The inbox holds seven candidate-lesson messages for this plan (001-007) written by lessons-capture and the retrospective. diff --git a/.plan/orchestrator/libraries-initial-implementation/landings/PLAN-21.md b/.plan/orchestrator/libraries-initial-implementation/landings/PLAN-21.md new file mode 100644 index 0000000..e49866b --- /dev/null +++ b/.plan/orchestrator/libraries-initial-implementation/landings/PLAN-21.md @@ -0,0 +1,102 @@ +# Landing Analysis: PLAN-21 — merge git-worktrees into git-core as a package, and adapt all documentation + +epic: libraries-initial-implementation +workstream: WS-01 +pr: #46 (https://github.com/cuioss/git-document-repository/pull/46), merged through the merge queue as `9e3cfb0` (squash, parent `dd57475`) + +> Landing record for one shipped plan. Written by the `analyze` verb from the inbox message `plan-21-merge-worktrees-into-core-008` +> after verifying it against git, the CI abstraction and the merged code. The message is a lead, never a fact. + +## Deliverable Fidelity vs Spec + +Diff read: `git diff -M dd57475 9e3cfb0` — 33 paths, +660/-935. The range `2075640..9e3cfb0` holds one more commit, `dd57475` (#45, the operator's: the application documentation `doc/app/**` moved to `cuioss/TrailHand`, plus two files of `.plan/project-architecture/` and part of the `AGENTS.md` and `doc/README.adoc` text), which does not belong to the plan. Every one of the 33 paths traces to a deliverable. `.github/project.yml` and every `pom.xml` dependency list are untouched; the root `pom.xml` only drops `git-worktrees`. +Of the spec's Expected Surface every touched entry is declared; the hypothesis entry `doc/adr/` was borne out (ADR-0001 and ADR-0004 changed, no new ADR, see below). Nothing undeclared was touched. + +| Deliverable (spec) | Verdict | Evidence | +|--------------------|---------|----------| +| 1. Move the code and tests into `git-core` | shipped-as-specified, with the log ids and the requirement ids decided at the outline | `git-core/src/main/java/module-info.java` 24-25 exports `de.cuioss.git.document.core` and `…core.worktrees`; four main classes and `package-info.java` (`@NullMarked`) and six test classes moved by `git mv` (renames at 93 to 99 % similarity; the only non-package source change is the S4449 null-check, below); `GitWorktreesLogMessages` merged into `GitCoreLogMessages` (INFO 2-4 `WORKTREE_ADDED`, `WORKTREE_REMOVED`, `BRANCH_RESERVATION_RELEASED`; WARN 102-104 `WORKTREE_CAPABILITY_REFUSED`, `REFLOG_NOT_WRITTEN`, `ROLLBACK_INCOMPLETE`; templates unchanged; no duplicate id, `GitCoreLogMessages.java` 56-127) and the catalogue in `git-core/doc/LogMessages.adoc` 20-55; `GitOutcome` Javadoc says "the worktrees package" | +| 2. Build configuration | shipped-as-specified | `git-worktrees/` and its POM removed; `git-core/pom.xml` dependency set unchanged, `consumer-isolation` lines 77-115 untouched, surefire `useModulePath=false` and the Lombok processor were already present and identical; `DomainTermAbsenceTest` 61-109 now fails when an exported package holds no scanned public type | +| 3. Specifications | shipped-modified (requirement ids renumbered at the operator's answer) | `git-core.adoc` 609-859 holds § Working Trees with the requirements as `GDR-CORE-12` to `GDR-CORE-15` (anchors at 650, 738, 776, 820; numbering `GDR-CORE-1..15` without a gap); `git-worktrees.adoc` deleted; `general.adoc` names the package in the summary, GDR-GEN-1, the suffix list, the GDR-GEN-3 table, GDR-NFR-3 and GDR-NFR-4 | +| 4. Concepts, decisions, client, index, root documents | shipped-modified (no new ADR: ADR-0004 rewritten in place and ADR-0001 adapted, at the operator's answer) | `module-structure.adoc`, the redrawn module graph SVG (six modules, three solid), `pm-mcp-core.adoc` (anchor renamed, xrefs follow), `doc/README.adoc`, `README.adoc`, `AGENTS.md` (table, Current state, the Enforcer sentence now names `git-store/pom.xml`); a repository content search for `git-worktrees`, `de.cuioss.git.document.worktrees`, `GDR-WT`, `GIT_WORKTREES` and `GitWorktreesLogMessages` outside `.plan/` finds nothing | + +The lane reports 4 deliverables and 4 done, as the spec. Not shown by the diff and recorded as unverified plan facts: the reactor test count 883 to 882 (one `@Test` is deleted, the duplicate `DomainTermAbsenceTest` of the removed module, so the number is plausible by reading), the green `verify`, `consumer-isolation` and `module-layering` passing (the PR merged through the merge queue under the required check `build / conclusion`, which is the corroborated CI fact). +**The owner decisions of the outline differ from the spec in three points**, all taken at the operator's answers relayed to the outline: the requirement ids were renumbered (the spec said they stay), the log ids INFO 2-4 / WARN 102-104 keep the worktrees refusal as the separate record `WORKTREE_CAPABILITY_REFUSED`, and ADR-0004 was rewritten in place with no new ADR (the spec said a new ADR). The spec's Done-when text "every moved test keeps its `GDR-WT-n` display name" and "the root pom lists `git-core` and `git-testing` only" (it predates `git-store`) therefore no longer describe the landing; the code and tests follow the decisions, and the three differences are info, not defects. The decision to rewrite an Accepted ADR in place carries no dated record of its reason in the ADR itself (info). +`merge_state=merged` and the merge commit `9e3cfb04b1a02246b826266b54f6095033ec80f2` are corroborated by `ci pr view` (state merged, base `main`). The queue squash message was not set by the run: the commit message on `main` carries the PR body including the CodeRabbit release-notes block (four hits for "coderabbit" in `git log -1 9e3cfb0`, as in `9bdc5ba`). + +**The four review-and-Sonar numbers were re-read from the provider.** The landing wrote `unresolved_review_threads`, `review_body_findings_total` and `review_body_findings_handled` as `unknown`; `runbook_lane review-gate --pr-number 46` shows 0 unresolved threads and 1 review body finding (CodeRabbit, inline-1) with no disposition recorded (`merge_allowed: false, reason: body_findings_undisposed`). The landing's own narrative says triage handled all four findings (three CodeRabbit, one Sourcery) and none was pending at merge; that is a lead for `review_body_findings_handled`, which only the run's progress file reproduces (a plan-marshall plan has none under the runbook parser), so it is **not re-verified**, and the single undisposed body finding the parser counts is not evidence that the landing lied. `sonar_open_new_issues=0`: the Sonar read for PR 46 returned `count_status: undecidable` (`head_revision_unavailable`, analysis revision `6b45edb`, the fix commit of the S4449 loop-back), so the number is **not re-verified**, and not a difference. + +## Three-Layer Analysis + +The landing was read on three layers by one read-only leaf, and the orchestrator re-read the points below at their file and line at the head (`9e3cfb0`). By reading: no build, no test run, no write. + +- **Dispatch:** one read-only `execution-context-level-3` leaf for this landing (resolved level `level-3`, `orchestrator.analyze` surface) — reason: a move of 10 files with 93 to 99 % renames, a log-id merge, a requirement renumbering and eleven documents is only checked by comparing the old and the new text, which is the depth the dispatch rule exists for, and one landing's read must not share context with the reconciliation of the seven lessons drained in the same scan. +- **Verdict:** layer 1 — all claims settled, none refuted, three differences between spec and owner decisions (info), one not shown; layer 2 — no blocker, no major, one minor (identical template under two log ids, a deliberate decision), residues info; layer 3 — no defect found. +- **Re-read by the orchestrator:** `git-core/src/main/java/module-info.java` 22-26; `GitCoreLogMessages.java` 51-126 (ids 1-4 and 100-104, templates); `JGitWorktreeOperations.java` 487-502 (the S4449 fix: `existing.getObjectId()` read once into `existingId`, the same two refusals, `StartPoint.of(existingId)`) and 962-971 (the `WORKTREE_CAPABILITY_REFUSED` log call, text and argument unchanged); `DomainTermAbsenceTest.java` 61-109; `git-core.adoc` anchors at 609 and 650 to 820; the repository-wide `git grep` for the old names at the head (empty outside `.plan/`); `AGENTS.md` module table 25-30; `git-core/pom.xml` 77-80; `.plan/project-architecture/` for the old module name (the check below). **Not re-read** (leaf-only, recorded as unverified leads): the per-line renaming diffs of the five moved test classes beyond the sampled `JGitWorktreeOperationsTest` display names, the SVG coordinates, the xref text of `pm-mcp-core.adoc` 104-216, and the JGit-side question of NFR-4 (whether a command the package drives reaches a translation bundle beyond `JGitText`). + +### Layer 1 — Specified + +| Finding | Claim | Verdict | Polarity | Evidence (file:line) | +|---------|-------|---------|----------|----------------------| +| L1-1 | Claims 0 to 8: the pre-plan observations (modules, file set, dependencies, duplicate test names, log-id collisions, `GitOutcome` text, imports, `.github`/`.claude` free of the name, ADR status) | checked | confirms | the leaf re-read each at `a558afd` / `dd57475`; claims 0 to 8 carry a corroborated verdict already, so the settlement is recorded as a duplicate and persists nothing | +| L1-2 | Claim 9 (hypothesis): the moved package has no reflection, resource or `ServiceLoader` access | checked | confirms | search of `git-core/.../core/worktrees/` empty for `Class.forName`, `getResource`, `ServiceLoader`, `ResourceBundle`, reflection, `MethodHandle`, `Proxy`, `getClassLoader`; the JGit-side half of GDR-NFR-4 is not shown (L1-4) | +| L1-3 | Claim 10 (hypothesis): the merged `git-core` still satisfies `consumer-isolation` | checked | confirms | dependency set and allow-list unchanged (`git-core/pom.xml` 14-66, 77-115); the build result is the merge under the required check | +| L1-4 | GDR-NFR-4: the JGit commands the package drives reach no further translation bundle than `JGitText` | not shown by reading | none | asserted in `general.adoc` 298-311, not shown; `NativeImageMetadataTest` was not run by the reading | +| L1-5 | The spec's Done-when and Standing Constraints against the owner decisions | checked | refutes (the text is superseded, not the claims) | `GDR-WT-n` display names and "no new ADR" are not what the spec says; see Fidelity. Info | + +Persisted onto the spec (`corpus set-verdict`): claims 9 and 10 corroborated at `9e3cfb0`. Claims 0 to 8 already carried a corroborated verdict, so no second write was made (equal strength, the first recorded stands). Claim 11 (the verify-first clause) is a procedure, not a claim. + +**Layer 1 verdict:** findings — 4 info; 1 not shown by reading (L1-4) + +### Layer 2 — Implemented + +| Finding | Severity | Claim | Evidence | +|---------|----------|-------|----------| +| L2-1 | minor | `CAPABILITY_REFUSED` (101) and `WORKTREE_CAPABILITY_REFUSED` (102) carry the identical template "Refused capability %s" under two ids, so a log consumer cannot tell them apart by text; deliberate (the owner kept the record separate and the spec says "texts stay"), not a defect | `GitCoreLogMessages.java` 96-110 | +| L2-2 | info | the Tests section of `git-core.adoc` still says "`AbsenceContractsTest` asserts …" for the core test while a second `AbsenceContractsTest` now lives in the worktrees package | `git-core.adoc` 861-873 | +| L2-3 | info | the `` of `git-core/pom.xml` reads awkwardly after "on JGit" | `git-core/pom.xml` 14 | +| L2-4 | minor | the committed architecture inventory `.plan/project-architecture/` still names the removed module: `_project.json` (2 hits), `git-document-repository/enriched.json` (1) and a whole `git-worktrees/enriched.json` (2) remain at the head. It is derived data (regenerated by the steward, not by this plan) and `.plan/` is outside the spec's content search, so no Done-when is broken; the two files #45 changed there are not part of this PR | `git grep` at `9e3cfb0`, `.plan/project-architecture/` | +| L2-5 | info | the reactor count 883 to 882, a green `verify`, `consumer-isolation`, `module-layering`, `NativeImageMetadataTest` and the Sonar read of 0 are not shown by reading | — | + +Every moved test keeps its assertions: only the package, the imports, the `DisplayName` requirement ids and the log record names changed (`JGitWorktreeOperationsTest` 372-1801; the INFO count by prefix at 1800 holds, the scenario logs no clone). + +**Layer 2 verdict:** findings — 2 minor, 3 info; the build and Sonar facts not shown by reading + +### Layer 3 — Correct + +No defect found in the moved code. The move is behaviour-neutral: the one non-package source change is the S4449 fix (`existingBranchStart`, 493-501: the same branch structure and outcomes, `NOT_FOUND` on a null id, one read of `getObjectId()`), and the one changed log call (`WORKTREE_CAPABILITY_REFUSED`, id 102) keeps its text and argument. Client-visible differences beyond the package: the artifact `git-worktrees` and its automatic module name are gone, `GitWorktreesLogMessages` and the prefix `GIT_WORKTREES` are gone (ids now `GIT_CORE-2..4` and `102..104`), and a consumer of `git-core` now gets the worktrees package exported; no result, outcome or message text changed, and nothing was released. The pre-existing layer-3 defects of PLAN-19 (L3-1 to L3-8) were moved unchanged and stay PLAN-20's deliverables 1 and 2; their claims were re-read and re-stamped at `9e3cfb0` in this act. + +**Layer 3 verdict:** findings — none; no blocker, no major, no minor + +### Conformance with the project's rules + +| Rule source | Rule | Verdict | Evidence | +|-------------|------|---------|----------| +| AGENTS.md | JPMS, `@NullMarked`, exports with the first public class of a package | conforms | `module-info.java` 24-25; `worktrees/package-info.java` keeps `@NullMarked` | +| AGENTS.md | expected outcomes are results, not thrown | conforms | `JGitWorktreeOperations.java` 962-971 unchanged in behaviour | +| AGENTS.md | `CuiLogger`, one `LogRecord` holder per module, a catalogue for every INFO/WARN/ERROR id | conforms | `GitCoreLogMessages` is the single holder; `git-core/doc/LogMessages.adoc` 20-55 matches the code | +| AGENTS.md | Javadoc on every public type and method; licence header | conforms | the new records carry Javadoc; moved files show no header change | +| AGENTS.md | no dependency added without approval; lower modules free of third-party dependencies | conforms | `git-core/pom.xml` dependency list identical; the allow-list untouched | +| AGENTS.md | a module adds its edge to `module-layering` in its own POM | conforms | `git-store/pom.xml` keeps its edge; `AGENTS.md` now points at it | +| AGENTS.md | do not touch `.github/project.yml` | conforms | absent from the diff | +| AGENTS.md / `general.adoc` | `DomainTermAbsenceTest` covers every exported package (GDR-GEN-1) | conforms | `DomainTermAbsenceTest.java` 61-109 asserts it and can fail | +| `general.adoc` | GDR-NFR-3: no secret in a log message | conforms | the moved templates carry a commit id or a refused capability only | +| `general.adoc` | GDR-NFR-4: the package has no reflection/resource access, so `git-core`'s metadata covers it | not shown by reading | own code clean (L1-2); the JGit side and `NativeImageMetadataTest` not run (L1-4) | +| Plan constraint | no assertion weakened or removed (one duplicate class deleted) | conforms | five moved test classes diffed | +| Plan constraint | Sonar open findings fixed or operator-selected | not shown by reading | the S4449 site is fixed (493-501); the Sonar read is undecidable on a merged PR | + +## Metrics and Anomalies + +- Tokens (landing fact `total_tokens`): 14,188,567 across the run; wall 10,740 s (about 3.0 h); model `claude-sonnet-5-5`, `effort` unknown; 13 finalize steps all `done`, one Sonar loop-back (iteration 1 of 3, TASK-006); branch cleanup `merged` through the merge queue. Against PLAN-07 (17.2 M tokens, 5.9 h) a documentation-heavy move with one loop-back cost four fifths of the tokens and half of the wall time. +- Cost drivers recorded by the lane: the outline re-entry that re-froze the success criteria on its own authority (the operator approved it afterwards; lesson `2026-11-20-07-001`), a Sonar `java:S4449` finding in a moved file that cost a whole loop-back (lesson `2026-11-18-07-007`), the simplify and quality-gate re-fires for a three-line commit (same lesson), and an automatic-review rate-window wait of about 35 minutes for a review that was not owed (lesson `2026-11-18-13-005`). +- Anomalies: the landing is incomplete by `landing-check` (4 keys: `effort`, `unresolved_review_threads`, `review_body_findings_total`, `review_body_findings_handled`; `cleanup_owed` and `sonar_open_new_issues` are now supplied), recorded as an Open Defect. The footprint-capture residue the retrospective names (10 rename sources read as unrealized, recall 76.7 %) is lesson `2026-11-20-17-001`; the commit trailer of `6b45edb` names "Claude Opus 5.5" while the session ran on Sonnet 5.5 (merged, not rewritten). + +## Routing / Merge Behavior + +Merged through the merge queue as a squash; the squash message carries the PR body (see Fidelity). The operator's #45 (`dd57475`, the application documentation moved out) merged first; the plan rebased onto it with no collision. Nothing else ran concurrently. PLAN-08, staged behind this plan by the operator's choice, shares three files with it (`AGENTS.md`, `doc/README.adoc`, `doc/specification/general.adoc`) and is now emittable. + +## Reconciliation Actions + +- Queue: PLAN-21 transitioned to `shipped`; `pr` 46, `landing` `landings/PLAN-21.md`, `plan_marshall_plan_id` `plan-21-merge-worktrees-into-core`. +- Spec corpus: PLAN-21 claims 9 and 10 corroborated at `9e3cfb0`; PLAN-20 re-grounded in the same act — the five worktrees claims (0 to 4) re-read and re-stamped at `9e3cfb0`, the worktrees paths, spec references, Expected Surface, Done-when base and sequencing text moved to the merged package, deliverable 2 extended with item (e) (L2-2, L2-3), deliverable 7 item (g) reduced to the note that the fourth domain-term copy is gone. +- Lessons (7 candidate messages): six are recurrences of existing lessons (`2026-11-20-07-001`, `2026-11-18-18-001`, `2026-11-18-07-007` twice, `2026-11-18-13-004` ninth plan, `2026-11-18-13-003`, `2026-11-18-13-005`) and one is new (`2026-11-20-17-001`, the realized footprint drops rename sources); a second new one is filed from the landing residue (the squash message carries the bot release-notes text). +- epic.md: the PLAN-21 entries (queue annotation, decision), the incomplete-landings entry extended to PLAN-21, a Watch for the stale `.plan/project-architecture` inventory (L2-4), the `GDR-WT` / `GIT_WORKTREES` names in the narrative marked as superseded. diff --git a/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-08-store-drafts-save.md b/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-08-store-drafts-save.md index 49982fa..ab49f1d 100644 --- a/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-08-store-drafts-save.md +++ b/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-08-store-drafts-save.md @@ -107,7 +107,7 @@ credential rules (GDR-STORE-8), with the concurrent-writers requirement GDR-NFR- - OBSERVED: `save.max-attempts` has no upper bound in `StoreConfiguration.readSave` (`git-store/src/main/java/de/cuioss/git/document/store/StoreConfiguration.java` 171-179: only `attempts < 1` is refused) and `ReferenceUpdater.swap` loops `attempt <= attempts` over an `int` (`git-core/src/main/java/de/cuioss/git/document/core/ReferenceUpdater.java` 199) — read at `a558afd`. - - verdict: corroborated | checked_at: a558afd517fda67e394ac4f249a5259a0007a305 | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: StoreConfiguration.java 171-179 refuses only attempts < 1; ReferenceUpdater.java 199 loops attempt <= attempts over an int; both re-read at a558afd + - verdict: corroborated | checked_at: 9e3cfb04b1a02246b826266b54f6095033ec80f2 | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: re-read at 9e3cfb0: StoreConfiguration.java 171-179 refuses only attempts < 1; ReferenceUpdater.java 199 loops attempt <= attempts over an int; neither file changed since a558afd (git diff a558afd 9e3cfb0 is empty for git-store and ReferenceUpdater.java) ## Expected Surface @@ -123,7 +123,8 @@ credential rules (GDR-STORE-8), with the concurrent-writers requirement GDR-NFR- - Depends on: PLAN-07 (module, registry, configuration, `PersonDirectory` interface for the author pseudonym; shipped, PR #41). - Runs after PLAN-21 (merge of `git-worktrees` into `git-core`; the operator's choice on 2026-11-20): the two share `AGENTS.md`, `doc/README.adoc` and `doc/specification/general.adoc`, and the merge rewrites the module table and the state lists this plan - extends. Emit it when PLAN-21 has landed. + extends. PLAN-21 has shipped (PR #46, `9e3cfb0`, 2026-11-20): the module table, the state lists and `general.adoc` already + describe the merged `git-core`, so PLAN-08 extends those texts and no `git-worktrees` module remains anywhere. - Overlaps with: every WS-02 plan — strictly sequential. - Adjacent to: GDR-STORE-12/13 (signature, trailers) — commits are written unsigned and without trailers here; PLAN-10 adds both on the same commit path. diff --git a/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-20-leftover-fixes-2.md b/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-20-leftover-fixes-2.md index 66f91b6..5b9d6bd 100644 --- a/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-20-leftover-fixes-2.md +++ b/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-20-leftover-fixes-2.md @@ -1,4 +1,4 @@ -# PLAN-20: Collected leftover fixes (git-worktrees, git-core, git-testing, build guards) +# PLAN-20: Collected leftover fixes (the worktrees package, git-core, git-store, git-testing, build guards) epic: libraries-initial-implementation workstream: WS-01 @@ -9,8 +9,8 @@ workstream: WS-01 ## Objective -Collect the leftovers the three-layer analyses of landed plans find — the layer-3 defects of PLAN-19 (`git-worktrees`) -and PLAN-04 (`git-core`), the minor layer-3 defects of PLAN-07 (`git-store`), the test gaps beside them, the `git-testing` leftovers of PLAN-18 and the residue of PLAN-06's build guards, which have no other +Collect the leftovers the three-layer analyses of landed plans find — the layer-3 defects of PLAN-19 (the worktrees code, +since PLAN-21 the package `de.cuioss.git.document.core.worktrees` of `git-core`) and PLAN-04 (`git-core`), the minor layer-3 defects of PLAN-07 (`git-store`), the test gaps beside them, the `git-testing` leftovers of PLAN-18 and the residue of PLAN-06's build guards, which have no other spec — and fix them in one plan instead of one plan per landing. This is the second collection plan, after PLAN-18. None of the items gives a wrong result on the default path: each needs a malformed configuration, a hostile or unusual input, or a failure in the middle of an operation. Later analyses fold further leftovers into this spec (see @@ -18,7 +18,7 @@ Dependencies and Sequencing) until the operator releases it. ## Deliverables -1. `git-worktrees` result contract (leftovers of PLAN-19's analysis, `landings/PLAN-19.md` L3-1 to L3-3 and L3-6): each +1. Worktrees package result contract (`git-core/src/main/java/de/cuioss/git/document/core/worktrees/`; leftovers of PLAN-19's analysis, `landings/PLAN-19.md` L3-1 to L3-3 and L3-6): each item is first reproduced by a test that fails at the base, then fixed. (a) `commit`, `worktreeStatus` and `worktreeRemove` end as a `FAILED` result with the class name only when JGit throws `IllegalArgumentException`, for example with `core.autocrlf = bogus` in the repository configuration; the reach is checked for every operation that @@ -30,18 +30,22 @@ Dependencies and Sequencing) until the operator releases it. raise, so a stored commit is not reported as thrown. - Skills: pm-dev-java:java-core, pm-dev-java:junit-core - Persona: implementer -2. `git-worktrees` paths, names and documentation (PLAN-19 analysis L3-4, L3-5, L3-7, L3-8 and L1-2): (a) `realPath` +2. Worktrees package paths, names and documentation (PLAN-19 analysis L3-4, L3-5, L3-7, L3-8 and L1-2; PLAN-21 analysis L2-1, L2-2): (a) `realPath` does not test existence and then resolve (the removal of a registered tree between the two calls ends `worktreeRemove` of an unrelated tree) and resolves a symlinked ancestor of a missing path, so a prunable tree removed through an alias is found; (b) the branch-name guard either applies git's rules for the names JGit 7.8.0 accepts and git refuses or reads differently (`HEAD`, a leading `-`, a component ending `.lock` inside the name, `@`, - the full `refs/heads/x` form), or the limit is recorded in `doc/specification/git-worktrees.adoc` with its reason + the full `refs/heads/x` form), or the limit is recorded in `doc/specification/git-core.adoc` § Working Trees (GDR-CORE-12 to GDR-CORE-15) with its reason — the choice follows what git itself accepts, which is checked against `git check-ref-format` in the compatibility test; (c) `worktreeAdd` on a bare root, which since PLAN-19 always ends `PATH_REFUSED` ("inside the git directory") while `git worktree add` works there: supported with a test, or the refusal documented in the Javadoc and the specification with a test that pins it; (d) the two stale Javadoc sentences (`WorktreeOperations` § `worktreeAdd` "until the new working tree is registered" and the `path` parameter, which omits the refusals for the git directory - and a linked root) and the two outcome lists of the specification that are still short. + and a linked root) and the two outcome lists of the specification that are still short; (e) two residues of PLAN-21's + landing (`landings/PLAN-21.md` L2-1, L2-2), both documentation: the Tests section of `doc/specification/git-core.adoc` + says "`AbsenceContractsTest` asserts ..." for the core test although a second `AbsenceContractsTest` now lives in the + worktrees package — it names both; and the `` of `git-core/pom.xml` reads awkwardly after "on JGit" — it + is reworded. - Skills: pm-dev-java:java-core, pm-dev-java:javadoc, pm-documents:ref-asciidoc - Persona: implementer 3. `git-core` write path (leftovers of PLAN-04's analysis, `landings/PLAN-04.md` L3-1 to L3-7, L2-4): each item is first @@ -59,7 +63,7 @@ Dependencies and Sequencing) until the operator releases it. a name conflict (`refs/heads/a/b` while `refs/heads/a` exists) from contention, running the operation once and not on every attempt, and a `refName` that is not a valid reference name (`refs/heads/../../../outside/x`) is refused before any file is touched by `compareAndSwap`, `compareAndDelete` and `compareAndSwapTree` (the notes entry points of - `GitNotes` already validate), with the outcome `git-worktrees` uses for an invalid name (`REFUSED_BY_POLICY`, no new + `GitNotes` already validate), with the outcome the worktrees package uses for an invalid name (`REFUSED_BY_POLICY`, no new constant), and a notes reference that name-conflicts with an existing one (`refs/notes/a/b` while `refs/notes/a` exists) is reported as that through `GitNotes.write`, not as a lost race; (e) edits that overlap as a file and a directory below it (`Put("a")` with `Put("a/b")`) are refused like a repeated path, or the loss is documented with a test; (f) `ReferenceUpdater.parseCommit` resolves with `parseAny` @@ -119,10 +123,10 @@ Dependencies and Sequencing) until the operator releases it. allow-list edit or the owner's approval: a rule or check bans every direct third-party dependency of `git-core` outside the approved list whatever its `optional` flag (the same rule with `searchTransitive=false` has no optional filter), shown by a deliberately wrong optional dependency; (b) the guards are proven by negative runs that are recorded only - in the squash message of PR #39 (`git-testing` at scope `compile` in `git-worktrees`; `cui-test-generator` at scope - `compile` in `git-core`): record the runs (the edit, the command, the failure message) in `general.adoc`, or make + in the squash message of PR #39 (`git-testing` at scope `compile` in `git-worktrees`, a module PLAN-21 has since removed, + so that run is repeated in `git-store` instead; `cui-test-generator` at scope `compile` in `git-core`): record the runs (the edit, the command, the failure message) in `general.adoc`, or make them repeatable by a check that needs no new library (a new library goes to the owner first); (c) the trade-offs the - review raised and the lane accepted (the four domain terms copied into three module tests, the allowed-edge + review raised and the lane accepted (the four domain terms copied into the module tests of `git-core`, `git-testing` and `git-store`, the allowed-edge includes that mirror `module-structure.adoc` by hand, a module widening its own edges in its own POM) are recorded as accepted exceptions with their reasons in `general.adoc`, or derived; (d) `NativeImageMetadataTest` also asserts the top-level shape of each file (an array of objects for `reflect-config.json`, an object for `resource-config.json`) and @@ -150,9 +154,9 @@ Dependencies and Sequencing) until the operator releases it. can fail it: either give the contract code to check (a verifying helper or a test-kit class other modules reuse) or record it as a documented contract test; (f) provider selection is tested on the class path only (`useModulePath=false`, the registrations in `META-INF/services`), so lookup in module mode (`provides`) is not exercised: add one module-path - test run or record the limit in `git-store.adoc`; (g) the four domain terms are now copied into four `DomainTermAbsenceTest` - classes (`git-core`, `git-worktrees`, `git-testing`, `git-store`): item (c) of deliverable 6 covers the three, extend it - to the fourth. + test run or record the limit in `git-store.adoc`; (g) the four domain terms are copied into the `DomainTermAbsenceTest` + classes of `git-core`, `git-testing` and `git-store` (PLAN-21 removed the fourth copy with `git-worktrees`): item (c) of + deliverable 6 names all three, so nothing remains here but that `git-store` is among them. - Skills: pm-dev-java:java-core, pm-dev-java:junit-core, pm-dev-java:java-security - Persona: implementer @@ -195,9 +199,9 @@ Dependencies and Sequencing) until the operator releases it. ## Done when - `verify` passes on the reactor at the plan's head, and every behaviour fix of deliverables 1 to 4 and of items (a), (b) and (c) of deliverable 7 has a test that - fails at `e43c7e915a9d0667117b72970bc156bbe73b6506` (`main` after PLAN-06, which changed no `git-core` or `git-worktrees` - code, so every defect below is present as at `14269fe`; the `git-worktrees` ones are present since `e2aced9`) and - passes at the plan's head. + fails at `9e3cfb04b1a02246b826266b54f6095033ec80f2` (`main` after PLAN-21, re-read at outline: the worktrees defects are + present since `e2aced9` and were moved by PLAN-21 without a behaviour change, the `git-core` ones are those the analyses of + PLAN-04 and PLAN-05 name at `14269fe`, the `git-store` ones those of `a558afd`) and passes at the plan's head. - Every item of deliverables 1 to 7 is either fixed and tested at the plan's head, or recorded as an accepted exception or an owner decision with its reason in the specification document of its module; no document lists an outcome the code does not return. @@ -207,24 +211,25 @@ Dependencies and Sequencing) until the operator releases it. ## Claim Labels - OBSERVED: `IllegalArgumentException` is caught by `worktreeAdd` and `worktreeList` but not by `commit`, `worktreeStatus` - and `worktreeRemove` — read at `git-worktrees/src/main/java/de/cuioss/git/document/worktrees/JGitWorktreeOperations.java` - § `commit` (742), `worktreeStatus` (854), `worktreeRemove` (704) at `e2aced9` (unchanged at `c8feb0b`); JGit 7.8.0 + and `worktreeRemove` — read at `git-core/src/main/java/de/cuioss/git/document/core/worktrees/JGitWorktreeOperations.java` + § `commit` (711, catch 743), `worktreeStatus` (840, catch 855), `worktreeRemove` (655, catch 705) against `worktreeAdd` + (catch 162) and `worktreeList` (catch 649) at `9e3cfb0` (the same code at `git-worktrees/.../worktrees/` at `e2aced9`); JGit 7.8.0 `Config.getEnum` throws for an unmatched enum value and `WorkingTreeOptions` reads `core.autocrlf` through it. - - verdict: corroborated | checked_at: c8feb0bb6af2baaa79f66736b8f9d5b6f591fe8c | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: re-read at e2aced9 in the PLAN-19 analysis; git diff e2aced9 c8feb0b -- git-worktrees is empty, so unchanged at c8feb0b; JGit 7.8.0 Config.getEnum / WorkingTreeOptions chain re-read there + - verdict: corroborated | checked_at: 9e3cfb04b1a02246b826266b54f6095033ec80f2 | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: re-grounded at 9e3cfb0 after PLAN-21 moved the code: JGitWorktreeOperations (git-core/.../core/worktrees) worktreeAdd catch 162 and worktreeList catch 649 carry IllegalArgumentException; commit catch 743, worktreeStatus catch 855 and worktreeRemove catch 705 do not; re-read by the orchestrator - OBSERVED: `releaseTarget` lets an `UncheckedIOException` from `Files.list` escape and `attempt` catches `IOException` only, so `rollBack` stops before `deleteCreatedBranch` — read at `JGitWorktreeOperations.java` § `releaseTarget` - (381-382), `attempt` (457-468), `rollBack` (433-447). - - verdict: corroborated | checked_at: c8feb0bb6af2baaa79f66736b8f9d5b6f591fe8c | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: re-read at e2aced9 in the PLAN-19 analysis; git diff e2aced9 c8feb0b -- git-worktrees is empty, so unchanged at c8feb0b + (372-385, `Files.list` at 381), `attempt` (457-468), `rollBack` (433-447) at `9e3cfb0`. + - verdict: corroborated | checked_at: 9e3cfb04b1a02246b826266b54f6095033ec80f2 | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: re-read at 9e3cfb0: releaseTarget 372-385 (Files.list at 381), attempt 457-468 catches IOException only, rollBack 433-447 runs releaseTarget before deleteCreatedBranch; the code moved without a behaviour change - OBSERVED: the success path calls `releaseReservation()` at 292, after the working tree is complete, and an `IOException` there reaches the add's catch, whose `rollBack` deletes the finished target — read at - `JGitWorktreeOperations.java` § `run` (292) and § `releaseReservation` (363-369). - - verdict: corroborated | checked_at: c8feb0bb6af2baaa79f66736b8f9d5b6f591fe8c | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: re-read at e2aced9 in the PLAN-19 analysis; git diff e2aced9 c8feb0b -- git-worktrees is empty, so unchanged at c8feb0b + `JGitWorktreeOperations.java` § `run` (292) and § `releaseReservation` (363-369) at `9e3cfb0`. + - verdict: corroborated | checked_at: 9e3cfb04b1a02246b826266b54f6095033ec80f2 | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: re-read at 9e3cfb0: run calls releaseReservation() at 292 after the working tree is complete; releaseReservation 363-369; unchanged by the move - OBSERVED: `realPath` is `Files.exists(path) ? path.toRealPath() : path.toAbsolutePath().normalize()`, so a missing path - does not resolve a symlinked ancestor and the check races — read at `JGitWorktreeOperations.java` § `realPath` (943-945). - - verdict: corroborated | checked_at: c8feb0bb6af2baaa79f66736b8f9d5b6f591fe8c | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: re-read at e2aced9 in the PLAN-19 analysis; git diff e2aced9 c8feb0b -- git-worktrees is empty, so unchanged at c8feb0b + does not resolve a symlinked ancestor and the check races — read at `JGitWorktreeOperations.java` § `realPath` (944-946) at `9e3cfb0`. + - verdict: corroborated | checked_at: 9e3cfb04b1a02246b826266b54f6095033ec80f2 | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: re-read at 9e3cfb0: realPath 944-946 is Files.exists(path) ? path.toRealPath() : path.toAbsolutePath().normalize() - OBSERVED: the reflog step of `commit` catches `IOException | IllegalArgumentException` only — read at - `JGitWorktreeOperations.java` § `commit` (758). - - verdict: corroborated | checked_at: c8feb0bb6af2baaa79f66736b8f9d5b6f591fe8c | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: re-read at e2aced9 in the PLAN-19 analysis; git diff e2aced9 c8feb0b -- git-worktrees is empty, so unchanged at c8feb0b + `JGitWorktreeOperations.java` § `writeLinkedReflogs` (755-762, catch 759) at `9e3cfb0`, called from `commit` (711). + - verdict: corroborated | checked_at: 9e3cfb04b1a02246b826266b54f6095033ec80f2 | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: re-read at 9e3cfb0: writeLinkedReflogs 755-762 catches IOException | IllegalArgumentException only (catch at 759), called from commit 711 - HYPOTHESIS: `Repository.isValidRefName` of JGit 7.8.0 accepts `HEAD`, a leading `-`, `a.lock/b` and `@` where git refuses or reads them differently — confirm/refute at JGit 7.8.0 `org.eclipse.jgit/src/org/eclipse/jgit/lib/Repository.java` § `isValidRefName` and `git check-ref-format` (verify-at-outline). @@ -335,8 +340,8 @@ Dependencies and Sequencing) until the operator releases it. ## Expected Surface -- OBSERVED: `git-worktrees/src/main/java/de/cuioss/git/document/worktrees/` -- OBSERVED: `git-worktrees/src/test/java/de/cuioss/git/document/worktrees/` +- OBSERVED: `git-core/src/main/java/de/cuioss/git/document/core/worktrees/` +- OBSERVED: `git-core/src/test/java/de/cuioss/git/document/core/worktrees/` - OBSERVED: `git-core/src/main/java/de/cuioss/git/document/core/` - OBSERVED: `git-core/src/test/java/de/cuioss/git/document/core/` - OBSERVED: `git-testing/src/main/java/de/cuioss/git/document/testing/` @@ -349,29 +354,30 @@ Dependencies and Sequencing) until the operator releases it. - OBSERVED: `git-store/pom.xml` - OBSERVED: `doc/specification/git-store.adoc` - HYPOTHESIS: `README.adoc` (the native-image wording, if it changes) -- OBSERVED: `doc/specification/git-worktrees.adoc` -- OBSERVED: `doc/specification/git-core.adoc` +- OBSERVED: `doc/specification/git-core.adoc` (§ Working Trees and the Tests section included) - OBSERVED: `doc/specification/git-testing.adoc` - HYPOTHESIS: `doc/specification/general.adoc` (GDR-GEN-3 table, if an outcome changes) - HYPOTHESIS: `doc/clients/pm-mcp-core.adoc` (the outcomes of the worktree operations, if one changes) - HYPOTHESIS: `AGENTS.md` ("Current state", if a public type or method changes) -- HYPOTHESIS: `git-worktrees/doc/LogMessages.adoc` and `git-core/doc/LogMessages.adoc` (only if a log message is added) +- HYPOTHESIS: `git-core/doc/LogMessages.adoc` (only if a log message is added) ## Dependencies and Sequencing - Seven deliverables, over the scope-bloat guard (about six): it proceeds unsplit because the operator asked for one collection plan, each deliverable is independent and touches its own module, and deliverable 7 was added at the operator's word on - 2026-11-20 (PLAN-07's landing). The operator may split it when releasing it; the natural cut is by module (deliverables - 1, 2 and the `git-worktrees` part of 6 move with PLAN-21; 3 and 4 `git-core`; 5 `git-testing`; 7 `git-store`). + 2026-11-20 (PLAN-07's landing). The operator may split it when releasing it; the natural cut is by package and module + (deliverables 1 and 2 the worktrees package of `git-core`; 3 and 4 the rest of `git-core`; 5 `git-testing`; 6 the build + guards; 7 `git-store`). - Collection plan: **hold it back from `next` until the operator releases it.** `analyze` folds further leftovers into this spec as new deliverables or as items of a deliverable, in the same edit as their Expected Surface; the operator releases the plan by naming it in a `next` round. It is queued at the end of WS-01. -- Depends on PLAN-21 (merge of `git-worktrees` into `git-core`, staged 2026-11-20, owner's call): after it the worktrees code, - its tests and its specification part live in `git-core` (`git-core/src/main/java/de/cuioss/git/document/core/worktrees/`, - `git-core.adoc` § Working Trees), `git-worktrees/` is gone and the module-layering include and the domain-term test of - that module with it. The paths, module names and surface entries of deliverables 1 and 2 and of the claims that name - `git-worktrees/` are re-grounded by the orchestrator when PLAN-21 lands, in the same act that stamps their verdicts; do - not emit this plan before that. +- Depends on PLAN-21 (merge of `git-worktrees` into `git-core`; shipped, PR #46, `9e3cfb0`, 2026-11-20): the worktrees code, its + tests and its specification part live in `git-core` (`git-core/src/main/java/de/cuioss/git/document/core/worktrees/`, + `git-core.adoc` § Working Trees, requirements GDR-CORE-12 to GDR-CORE-15, log ids GIT_CORE-2 to 4 and 102 to 104), + `git-worktrees/` is gone and the module-layering include and the domain-term test of that module with it. The paths, + names, surface entries and claims of deliverables 1, 2, 3 and 6 and of the Done-when base were re-grounded at `9e3cfb0` in + the act that landed PLAN-21's analysis (the five worktrees claims stamped there); the owner decisions of that plan's outline + (ids renumbered, ADR-0004 rewritten in place) are in force. - Also depends on: PLAN-04 (landed as PR #35, merge commit `c8feb0b`), PLAN-05 (PR #37, `14269fe`), PLAN-06 (PR #39, `e43c7e9`), PLAN-19 (PR #31, `e2aced9`) and PLAN-18 (PR #28). - Overlaps with: nothing in flight. PLAN-05 and PLAN-06, which shared entries with it, have shipped. It shares the root diff --git a/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-21-merge-worktrees-into-core.md b/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-21-merge-worktrees-into-core.md index 3f85d6e..96eae78 100644 --- a/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-21-merge-worktrees-into-core.md +++ b/.plan/orchestrator/libraries-initial-implementation/plans/PLAN-21-merge-worktrees-into-core.md @@ -175,8 +175,10 @@ package or its specification is adapted in the same plan. native-image metadata covers the merged package — confirm/refute by a search of the moved package for `Class.forName`, `getResource`, `ServiceLoader` and `ResourceBundle` (verify-at-outline; the PLAN-06 analysis searched the main sources of both modules and found none). + - verdict: corroborated | checked_at: 9e3cfb04b1a02246b826266b54f6095033ec80f2 | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: three-layer leaf grep of git-core/.../core/worktrees/ at 9e3cfb0 finds no Class.forName, getResource, ServiceLoader, ResourceBundle, reflection, MethodHandle, Proxy or getClassLoader; the JGit side (a translation bundle beyond JGitText reached through CommitCommand, DirCacheCheckout, Status, hooks) is not shown by reading and NativeImageMetadataTest was not run by the reading - HYPOTHESIS: the merged `git-core` still satisfies the `consumer-isolation` allow-list, because the moved code needs no artifact `git-core` does not already declare — confirm/refute by the `verify` run of the reactor (verify-at-outline). + - verdict: corroborated | checked_at: 9e3cfb04b1a02246b826266b54f6095033ec80f2 | by: libraries-initial-implementation/analyze | rescoped: n/a | evidence: git-core/pom.xml dependency set unchanged and consumer-isolation allow-list lines 77-115 untouched at 9e3cfb0 (leaf reading, re-read by the orchestrator); PR #46 merged through the merge queue under the required check build / conclusion - Verify-first clause: the numbers above are those of `ed2abde`; every OBSERVED claim is re-read at the plan's own HEAD at outline, and a refuted one is dropped from the deliverable that names it. The set of documents that name the module is derived again by a content search at the plan's HEAD, not taken from this list. diff --git a/.plan/orchestrator/libraries-initial-implementation/queue-view.md b/.plan/orchestrator/libraries-initial-implementation/queue-view.md index 3574f5d..81df1c7 100644 --- a/.plan/orchestrator/libraries-initial-implementation/queue-view.md +++ b/.plan/orchestrator/libraries-initial-implementation/queue-view.md @@ -4,7 +4,7 @@ ## START HERE -**Resume anchor**: PLAN-07 (git-store) shipped (PR #41, 9bdc5ba); nothing is running. PLAN-21 (merge git-worktrees into git-core, all documentation adapted) is emitted and awaits the operator's launch and then 'running'; its package is decided, requirement ids / log ids / ADR form are asked at its outline. PLAN-08 is staged and runs after PLAN-21 (three shared files; it carries the save.max-attempts bound from PLAN-07's analysis). PLAN-20 (seven deliverables, held back) depends on PLAN-21 and is re-grounded when PLAN-21 lands. Open question to the operator: the 0.1.0 release (git-store is now on main and would be published in it; PLAN-21 and PLAN-20 before it?). Ledger changes since PR #40 are landed only on the operator's word. +**Resume anchor**: PLAN-21 (git-worktrees merged into git-core, PR #46, 9e3cfb0) shipped and analysed; nothing is running. PLAN-08 (git-store drafts, explicit save, expiry; carries the save.max-attempts bound) is emitted and awaits the operator's launch and then 'running'; PLAN-09..11 and PLAN-12..17 are behind it. PLAN-20 (seven deliverables, re-grounded at 9e3cfb0, claims 0-4 re-stamped) stays held back until the operator names it in a next round. Open: the 0.1.0 release question (git-store is on main; PLAN-20 and the git-store plans against the cut); the stale .plan/project-architecture inventory (Watch); person-bound draft reference names re-checked at PLAN-08's outline; incomplete landings PLAN-01..07, 18, 19, 21. Ledger changes since the land of PR #44 are landed only on the operator's word. **Phase**: orchestrating **Queue** (staged, in order): 1. PLAN-08 (WS-02) @@ -18,7 +18,6 @@ 9. PLAN-16 (WS-04) 10. PLAN-17 (WS-05) 11. PLAN-20 (WS-01) -12. PLAN-21 (WS-01) - PLAN-01 (WS-01) — plan=git-testing-fixtures — PR 23 — landing=landings/PLAN-01.md — status: shipped - PLAN-02 (WS-01) — plan=git-core-import — PR 25 — landing=landings/PLAN-02.md — status: shipped - PLAN-03 (WS-01) — plan=plan-03-git-worktrees-import — PR 30 — landing=landings/PLAN-03.md — status: shipped @@ -28,6 +27,7 @@ - PLAN-07 (WS-02) — plan=plan-07-store-module-registry-spi — PR 41 — landing=landings/PLAN-07.md — status: shipped - PLAN-18 (WS-01) — plan=plan-18-leftover-fixes — PR 28 — landing=landings/PLAN-18.md — status: shipped - PLAN-19 (WS-01) — plan=plan-19-git-worktrees-hardening — PR 31 — landing=landings/PLAN-19.md — status: shipped +- PLAN-21 (WS-01) — plan=plan-21-merge-worktrees-into-core — PR 46 — landing=landings/PLAN-21.md — status: shipped ## Ordered Queue @@ -43,5 +43,4 @@ | 8 | PLAN-15 | WS-04 | staged | AGENTS.md; README.adoc; doc/README.adoc; doc/concepts/module-structure.adoc; doc/specification/git-directory.adoc; pom.xml | plan-marshall | | 9 | PLAN-16 | WS-04 | staged | AGENTS.md; doc/README.adoc; doc/concepts/module-structure.adoc; doc/specification/git-directory.adoc | plan-marshall | | 10 | PLAN-17 | WS-05 | staged | AGENTS.md; README.adoc; doc/README.adoc; doc/concepts/module-structure.adoc; doc/specification/general.adoc; doc/specification/git-timestamp.adoc; pom.xml | plan-marshall | -| 11 | PLAN-20 | WS-01 | staged | AGENTS.md; README.adoc; doc/clients/pm-mcp-core.adoc; doc/specification/general.adoc; doc/specification/git-core.adoc; doc/specification/git-store.adoc; doc/specification/git-testing.adoc; doc/specification/git-worktrees.adoc; git-core/doc/LogMessages.adoc; git-core/pom.xml; git-core/src/main/java/de/cuioss/git/document/core/; git-core/src/main/resources/META-INF/native-image/; git-core/src/test/java/de/cuioss/git/document/core/; git-store/pom.xml; git-store/src/main/java/de/cuioss/git/document/store/; git-store/src/test/java/de/cuioss/git/document/store/; git-testing/src/main/java/de/cuioss/git/document/testing/; git-testing/src/test/java/de/cuioss/git/document/testing/; git-worktrees/doc/LogMessages.adoc; git-worktrees/src/main/java/de/cuioss/git/document/worktrees/; git-worktrees/src/test/java/de/cuioss/git/document/worktrees/; pom.xml | plan-marshall | -| 12 | PLAN-21 | WS-01 | staged | AGENTS.md; README.adoc; doc/README.adoc; doc/adr/; doc/clients/pm-mcp-core.adoc; doc/concepts/module-structure.adoc; doc/resources/diagrams/module-dependency-graph.svg; doc/specification/general.adoc; doc/specification/git-core.adoc; doc/specification/git-worktrees.adoc; git-core/doc/LogMessages.adoc; git-core/pom.xml; git-core/src/main/java/; git-core/src/test/java/; git-worktrees/; pom.xml | plan-marshall | +| 11 | PLAN-20 | WS-01 | staged | AGENTS.md; README.adoc; doc/clients/pm-mcp-core.adoc; doc/specification/general.adoc; doc/specification/git-core.adoc; doc/specification/git-store.adoc; doc/specification/git-testing.adoc; git-core/doc/LogMessages.adoc; git-core/pom.xml; git-core/src/main/java/de/cuioss/git/document/core/; git-core/src/main/java/de/cuioss/git/document/core/worktrees/; git-core/src/main/resources/META-INF/native-image/; git-core/src/test/java/de/cuioss/git/document/core/; git-core/src/test/java/de/cuioss/git/document/core/worktrees/; git-store/pom.xml; git-store/src/main/java/de/cuioss/git/document/store/; git-store/src/test/java/de/cuioss/git/document/store/; git-testing/src/main/java/de/cuioss/git/document/testing/; git-testing/src/test/java/de/cuioss/git/document/testing/; pom.xml | plan-marshall | diff --git a/.plan/orchestrator/libraries-initial-implementation/queue/PLAN-21.json b/.plan/orchestrator/libraries-initial-implementation/queue/PLAN-21.json index 6e404bf..b8e1b4d 100644 --- a/.plan/orchestrator/libraries-initial-implementation/queue/PLAN-21.json +++ b/.plan/orchestrator/libraries-initial-implementation/queue/PLAN-21.json @@ -2,10 +2,10 @@ "id": "PLAN-21", "slug": "merge-worktrees-into-core", "workstream": "WS-01", - "status": "staged", + "status": "shipped", "lane": "plan-marshall", - "plan_marshall_plan_id": "", - "pr": "", - "landing": "", + "plan_marshall_plan_id": "plan-21-merge-worktrees-into-core", + "pr": "46", + "landing": "landings/PLAN-21.md", "seq": 21 } \ No newline at end of file diff --git a/.plan/orchestrator/libraries-initial-implementation/resume_anchor.md b/.plan/orchestrator/libraries-initial-implementation/resume_anchor.md index 6d145f9..cb84384 100644 --- a/.plan/orchestrator/libraries-initial-implementation/resume_anchor.md +++ b/.plan/orchestrator/libraries-initial-implementation/resume_anchor.md @@ -1 +1 @@ -PLAN-07 (git-store) shipped (PR #41, 9bdc5ba); nothing is running. PLAN-21 (merge git-worktrees into git-core, all documentation adapted) is emitted and awaits the operator's launch and then 'running'; its package is decided, requirement ids / log ids / ADR form are asked at its outline. PLAN-08 is staged and runs after PLAN-21 (three shared files; it carries the save.max-attempts bound from PLAN-07's analysis). PLAN-20 (seven deliverables, held back) depends on PLAN-21 and is re-grounded when PLAN-21 lands. Open question to the operator: the 0.1.0 release (git-store is now on main and would be published in it; PLAN-21 and PLAN-20 before it?). Ledger changes since PR #40 are landed only on the operator's word. +PLAN-21 (git-worktrees merged into git-core, PR #46, 9e3cfb0) shipped and analysed; nothing is running. PLAN-08 (git-store drafts, explicit save, expiry; carries the save.max-attempts bound) is emitted and awaits the operator's launch and then 'running'; PLAN-09..11 and PLAN-12..17 are behind it. PLAN-20 (seven deliverables, re-grounded at 9e3cfb0, claims 0-4 re-stamped) stays held back until the operator names it in a next round. Open: the 0.1.0 release question (git-store is on main; PLAN-20 and the git-store plans against the cut); the stale .plan/project-architecture inventory (Watch); person-bound draft reference names re-checked at PLAN-08's outline; incomplete landings PLAN-01..07, 18, 19, 21. Ledger changes since the land of PR #44 are landed only on the operator's word.