chore(issues): archive closed issue #2517 spec

[josecelano]

Archives the specification for #2517 (closed on GitHub after PR #2527 merged) from docs/issues/open/ to docs/issues/closed/.

Related to #2517. Part of EPIC #1669.

• Verified Make the statistics repositories' metrics lock private #2517 is CLOSED on GitHub and PR refactor: [#2517] make the statistics repositories' metrics lock private #2527 is merged.
• Moved docs/issues/open/2517-1669-encapsulate-statistics-repository-state/ (spec, manual verification evidence, review reports, and evidence/ artifacts) to docs/issues/closed/.
• Updated the spec frontmatter (status: done, spec-path, last-updated-utc), ticked the final workflow checkpoint, and added a progress-log entry; updated the issue-spec path in manual-verification-evidence.md.
• Updated EPIC Overhaul: Packages #1669: ticked the quick-list entry, marked the subissue row DONE with the closed path, updated the specs index link, and added a progress-log entry.
• Searched the repository for the folder name; no stale docs/issues/open/ references remain.
• Pre-commit and pre-push checks passed.

The specs for #2493 and #2505 are also closed on GitHub but are intentionally not archived here.

[Copilot]

Warning

Copilot couldn't run its full agentic review because it didn't start before the timeout. Make sure your repository has a runner available, or add a copilot-code-review.yml file specifying one with the runs-on attribute. See the docs for more details.

Copilot review overview

0 open findings

What changed in this PR

Archives the specification for issue #2517 by updating spec metadata/links and marking completion across the issue spec and its parent EPIC.

Changes:

• Updated #2517 spec metadata to status: done, moved spec-path to docs/issues/closed/, and checked the final workflow checkpoint.
• Updated manual verification evidence to point at the closed spec path.
• Updated EPIC #1669 to mark #2517 as DONE and repointed index/table links to the closed spec path.

File	Description
docs/​issues/​open/​2517-1669-encapsulate-statistics-repository-state/​manual-verification-evidence.md	Updated issue-spec link and timestamp to reference the closed spec path.
docs/​issues/​open/​2517-1669-encapsulate-statistics-repository-state/​ISSUE.md	Marked the issue spec as done; updated spec-path, timestamp, checklist, and progress log.
docs/​issues/​open/​1669-overhaul-packages/​EPIC.md	Marked #2517 as DONE and updated links/table rows to point at the closed spec.

🧠 Review effort: Lite

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[da2ce7]

Reviewed at 87dec325aab8ec7c14a956751208e7858c425e7e (round 1 of the loop's review, posted at the rebased head; the round computed at 77e718306 was superseded by the rebase onto 988c3bc76 before it posted). Recomputed from the bytes at this head, against the merge base 988c3bc76 (= develop).

Rebase neutrality. git range-diff 184446a2b..77e718306 988c3bc76..87dec325a shows one = row. After normalising index lines, the two patch texts (-U3 and -U0) differ only in two ISSUE.md hunk headers, which move up one line (@@ -214,7 +214,7 @@ → -213,7 +213,7, @@ -296,6 +296,8 @@ → -295,6 +295,8) because #2528 replaced three Out of Scope lines with two. The similarities (R098, five R100, R098, M), seven of the eight result blobs, the message, the author and the author time are unchanged. The head's tree equals the round-1 trial merge of 77e718306 with 988c3bc76. The archived ISSUE.md carries "tracked by #2521" (:131-133, from 8901bc508), and nothing remains under docs/issues/open/2517-….

Move. Seven files are renamed, five of them byte-identical. ISSUE.md gets status: done, the closed/ spec-path, stamp 14:31 → 15:06, checkpoint :216 ticked and log entry :298-299. The evidence record's issue-spec and stamp (14:15 → 15:06) move together. Step 4's rule, "set last-updated-utc to the current date when that field exists", holds: each 15:06 stamp postdates every earlier edit of its file, including 8901bc508 (14:51:05Z) in ISSUE.md. The ticked checkpoint is true: #2517 closed at 15:04:04Z after 184446a2b carried Closes #2517.

References. docs/issues/open/2517 and open/2517 have no hits at this head. Of the nine folder-name hits, five are the new closed/ paths and four the branch name (ISSUE.md:10, :229; agent-review-reports.md:19, :131). Every Markdown link and frontmatter path in the four touched documents resolves.

What the rebase settled. #2528's Out of Scope edit is now in the base and moves with the rename, so no merge question remains. The capture reads mergeable: true, mergeable_state: unstable, with base.sha at 988c3bc76.

EPIC. The EPIC blob is identical to round 1's. Quick list :671 reads "— DONE (2026-10-14, PR #2527)". Details row :735 is DONE with ../../closed/…, like the #2446 row. Specs Index :758 is edited in place, and log entry :1016-1017 has the form of 3127250e6. The stamp goes 09:44 → 15:06.

Stale text. Of the loop's post-merge findings on PR #2527, F4 is closed by the evidence stamp bump and the Out of Scope half of F2 by 8901bc508. The Risks half of F2 now contradicts Out of Scope inside this head; it and F3 remain as inline F1 and F2, Suggestions.

Hygiene. The subject is Step 6's single-issue form, untagged like the earlier archive commits, with no prose body. Author and committer are one identity; the committer time 15:29:24Z (the rebase) follows the author time 15:07:53Z. One trailer line; git diff --check clean; the banned token absent. Related to #2517 fits a docs/issues/-only diff, and no closing keyword is in the title, body or message. The body is unchanged and each bullet still matches the diff.

Review bot. Its only review is the overview at 77e718306: no inline comments, no numbered findings, nothing to rule.

Findings. F1 and F2, both Suggestion; none blocking.

Checked, no finding. cleanup-completed-issues Steps 0-3 and 5 (Step 0.7 does not apply; the EPIC stays open). The title drops the template's "to docs/issues/closed" tail, which is harmless. The pre-commit and pre-push claim is left to the gate of record.

Checks

On the loop's compute hub at this head, merge base 988c3bc76 (= develop; receipt server-gates-krkavec-288): pre-commit profile gate exit 0 (76 s); linter markdown, linter cspell, linter lychee exit 0; frontmatter-validator --all: 19 errors / 4 warnings at this head and 19 / 4 at 988c3bc76 re-measured, none in the files this PR moves or edits. GitHub checks at this head at posting time: "Package Coverage (${{ matrix.package }}): completed/skipped; Package Coverage Regression: completed/success; Docker E2E: completed/skipped; Layer Boundary Bans: completed/success; Unit (stable): completed/success; Test (Docker) (release): queued/pending; Documented Clippy Allows: completed/success; Unit (nightly): completed/success; Generate Coverage Report: completed/success; Discover Package Coverage: completed/success; Docs Lint: completed/success; Build on ubuntu-latest (nightly): completed/success; Build on macos-latest (stable): completed/success; Build on windows-latest (nightly): completed/success; Build on windows-latest (stable): completed/success; Build on ubuntu-latest (stable): completed/success; Build on macos-latest (nightly): completed/success". At posting, develop is 988c3bc76.

[da2ce7]

ACK 87dec32 — verified the rebase is patch-neutral against 77e7183 (one equal range-diff row, hunk offsets only), the seven-file rename now carrying the tracked-by-#2521 bullet, the 15:06 stamps postdating every edit, and zero old-path references at this head

[josecelano]

ACK ab6ea7d

[da2ce7]

Reviewed at ab6ea7da65fa8c3883bb8be51275b48dc0fdd983 (post-merge). Recomputed from the bytes at the merged head, against the merge base 988c3bc76; the PR merged at 16:28:07Z as b24550171.

Label: POST_MERGE_FINDINGS — this round answers the standing review request on this PR, which merged before the round ran. The two findings below stand at the merged head for a follow-up; neither is blocking, and no ACK follows a merged PR.

Delta. Four commits sit on 87dec325a, the head the loop's round 2 approved: fa48afeb3, b97093fe9, c8cae256b and ab6ea7da6. They change the archived ISSUE.md (+14/−4) and manual-verification-evidence.md (+10/−5) and add two audits (+167, +124). The other eight paths are byte-identical to 87dec325a. b24550171 differs from the head on these paths only by #2529's EPIC row (:713), which came from develop.

The PR #2527 findings.

• F1: the correction is right in place and wording. The evidence (:20-23) annotates docs(statistics): [#2517] introduce the repository privacy doctests under the five refactor(<package>): [#2517] make the statistics repository lock private subjects it lists, and the log entry (:305) names both subjects exactly. The bytes agree: that commit adds only the two lead-in lines per file, and each refactor commit adds one compile_fail,E0616 block and one companion.
• F2: Risks (:373-374) reads "#2521 (see Out of Scope) touches the same udp-core and udp-server files", and Out of Scope (:131-133) reads "tracked by #2521". The only "not committed yet" left is the historical 09:06 entry (:236). The #2521 spec is still under open/ on develop.
• F3: :147-149 now say one copy of each is stored "except the stats-before reading", which rests on the diff -u transcript. evidence/ holds the four named files.
• F4: both archived stamps read 16:00; the last edit is c8cae256b at 16:00:25Z, and nothing touches the EPIC after the archive commit that set its 15:06 stamp.

The round-2 F1 and F2 are fixed by the same commits. The log entries 15:57, 15:59 and 16:00 each precede their commits (15:58:16Z, 15:59:04Z, 16:00:25Z).

Audits. The pr-2530-review record has two rows, F1 and F2, both Suggestion and FIXED/RESOLVED. Source and reply ids match the capture, and both threads are resolved there. Its validator run (--base 988c3bc76, capture comments) is ok: 2 rows, 2 log entries, 0 failures. The bot's overview correctly adds no row. The pr-2527-review record is new at this head: approval URL in Ownership, four FOLLOW_UP/OPEN rows (F1 Minor, F2-F4 Nit, matching the review), and current-tree line claims that all hold. Its threads, replies, the approval comment and review times are not in this capture. Run without them, the validator fails only the eight source and reply lookups; the row/detail and log-order checks pass. One gap in the PR #2530 record: F3.

Hygiene. The three fix subjects are tagged [#2517], like 8901bc508. The audit commit is untagged, like the PR #2528 audit (620b2a606). Each commit has no prose body and one trailer line, and author = committer in identity and time. git diff --check is clean and the banned token is absent. The PR body is byte-identical to round 1's: F4.

Findings. F3 and F4, both Nit; none blocking.

Checked, no finding.

• Ancestry: the merge base is 988c3bc76.
• Every link and frontmatter path in the touched spec files resolves.
• The audits' 16:01 log entries come about a minute before the replies they record (16:02:28Z-16:02:30Z).
• docs(pr-reviews) scope.
• The later close-out of the PR #2527 rows on develop (e2e21f20c) is context only.

Checks

On the loop's compute hub at the merged head, merge base 988c3bc76 (the first parent of the merge commit b24550171; receipt server-gates-krkavec-296): pre-commit profile gate exit 0 (83 s); linter markdown, linter cspell, linter lychee exit 0; frontmatter-validator --all: 19 errors / 4 warnings at this head and 19 / 4 at 988c3bc76 re-measured, none in the files this PR touches; validate-audit-record.py --pr-number 2530 exit 0 against this head's pr-2530-review record (2 rows, 2 log entries, 0 failures). GitHub checks at this head at posting time: "Publish (Release): completed/skipped; Publish (Development): completed/skipped; Context: completed/success; Package Coverage (${{ matrix.package }}): completed/skipped; Package Coverage Regression: completed/success; Docker E2E: completed/skipped; Unit (nightly): completed/success; Layer Boundary Bans: completed/success; Unit (stable): completed/success; Documented Clippy Allows: completed/success; Test (Docker) (release): completed/success; Build on ubuntu-latest (nightly): completed/success; Build on ubuntu-latest (stable): completed/success; Build on windows-latest (nightly): completed/success; Build on macos-latest (stable): completed/success; Build on windows-latest (stable): completed/success; Build on macos-latest (nightly): completed/success; Generate Coverage Report: completed/success; Discover Package Coverage: completed/success; Docs Lint: completed/success". At posting, develop is 48c4d3013.

[da2ce7]

[Nit][F3] The record keeps review 5520610065 as APPROVED, but GitHub dismissed it before the audit was committed

This line reads "APPROVED, submitted 2026-10-14 15:43:03 UTC". The API has the review DISMISSED, submitted 15:43:04Z (15:43:03Z is the inline comments' time). The timeline has review_dismissed at 16:01:46Z, when the F1-F2 fix commits were pushed, four minutes before this audit's commit (16:05:25Z). Neither the paragraph nor the 16:01 Processing Log entry records it. The repository's own records note stale dismissals: pr-2178-review lists such reviews as DISMISSED with the reason, and pr-2363-review logs "GitHub dismissed round 4 … as stale when … pushed". A follow-up could add one sentence and one log line. Non-blocking.

[da2ce7]

[Nit][F4] The PR description was not updated when this branch took on the PR #2527 follow-up

This is outside the diff and anchored here, where the record says it was created on this PR's branch. The PR body at merge is byte-identical to the round-1 body: it describes the archive alone and names neither the approved PR #2527 follow-up (F1-F4, the approval comment), nor the three fix commits, nor the two audits. process-pr-review step 4 asks to "link the follow-up PR to the merged PR", and open-pull-request asks the body to list the files touched. The merge tool copied that body into b24550171, so the merge commit cannot change now. Editing the PR description on GitHub, or adding one line to this record, would leave the follow-up findable from the PR. Non-blocking.