fix(docs): correct the 2026.299.0822 entry's false clause about the FAIL: marker by aitrawog · Pull Request #194 · opencharly/docs

aitrawog · 2026-10-26T08:30:54Z

Summary

Head 7317be448a1668e6d7f4d12e1a2565f7f37815f8 — 1 file changed, 2 insertions(+).

A correction to the published record, not a change to the product. CHANGELOG/2026.299.0822.md — the entry tag-on-merge wrote from #192's body at merge time — carries a sentence that is false, and it contradicts the sentence right after it:

Both harnesses now carry the marker on both failure paths, and the marker is asserted by a test that can fail — the harness itself, whose every mutant must turn the self-test red. Nothing asserts on these messages; they are read by a human.

The first clause is false. The second sentence says so.

The finding's origin, stated plainly. The charly/pr-validator raised it on #192's second review as a non-blocking note, and it is right:

The body's round-2 paragraph contains a small internal wobble: "the marker is asserted by a test that can fail — the harness itself…" is immediately followed by "Nothing asserts on these messages; they are read by a human." The second sentence is the true one; nothing greps FAIL:.

Nothing asserts on the FAIL: marker. The mutation harness asserts that a mutant turns the self-test red — on the case that covers the mechanism it removes — and it does not assert on the message text. No CI step greps for the prefix. The clause describes a test that does not exist.

The clause was written to say something true, and was phrased as something false. It was meant to say "this change is covered by a test that can fail" — which is true, and is the change's central property — but it was phrased as "the marker is asserted by a test". The sentence right after it was written from the correct understanding of the same fact, and the two were never reconciled.

The correction is APPENDED, and nothing is rewritten. The false sentence stays where it is; the correction sits immediately after it, where a reader meets it, and the two can be read beside each other.

file	what changes
`CHANGELOG/2026.299.0822.md`	one content line appended after the paragraph that carries the false clause — the correction, which names #193 in its own opening — plus the blank separator that follows it (`+2/-0` in total)

The merged PR body is NOT edited, and the merged entry's original text is NOT written over. The body is the artefact the verdict was produced against; the entry is the published history, and tag-on-merge wrote it from that body, so the two are the same text. Rewriting either in place would desynchronize the two surfaces tag-on-merge keeps identical, and would erase the record of what was validated. The precedent for exactly this shape is opencharly/opencharly#582: "it needs a follow-up commit that adds a correction, not a rewrite … the correction belongs in history where the divergence can be read beside it."

The code is not touched. The two harnesses are byte-identical to the merge — the diff touches zero lines under .github/scripts/ — and the validator reproduced #192's diff. The landing squash's own commit message states the truth already: "the exit code is the real signal -- nothing in CI greps for the prefix -- but two failure paths in one tool should read alike". Only the entry's prose carried the false clause.

The true statement, for the record. The FAIL: marker is a human-facing convention: it makes the two failure paths of one tool read alike in a CI log, and it is read by a person. It is not asserted by any test, and no CI step greps for it. What the mutation harness proves — and what #192 strengthened — is a different thing and remains proven: every mutant in each harness turns the self-test red, on the case that covers the mechanism it removes, five per harness, ten in total.

How tested

The change-class gate, on the committed tree — this repo's own gate, under the node this repo's CI uses (.node-version → 22.23.4 → npm 10.9.9):

=== node/npm ===
v22.23.4
10.9.9
=== git ===
7317be4 fix(docs): correct the 2026.299.0822 entry's false clause about the FAIL: marker
2	0	CHANGELOG/2026.299.0822.md
=== npm ci ===

added 306 packages in 2s
CI_RC=0

The advisory gate, its self-test, and its mutation harness — the repo's own gates, unchanged by this diff, re-run so the correction lands on a green tree:

### python3 .github/scripts/check_dependency_advisories.py
check-dependency-advisories: OK (no known advisory in the production tree or the dev tree)
RC=0
### python3 .github/scripts/check_dependency_advisories.py --self-test
check-dependency-advisories self-test: OK (a production finding, a dev finding and a clean tree are all told apart, and an unauditable tree is an error)
RC=0
### python3 .github/scripts/test_check_dependency_advisories.py
ok  mutA-noproductionaudit — the production audit removed, and the self-test turned red on it
ok  mutB-nofindings — the findings loop removed, and the self-test turned red on it
ok  mutC-invertedlabel — the tree label removed, and the self-test turned red on it
ok  mutD-swallowederror — the report guard removed, and the self-test turned red on it
ok  mutE-noadvisories — the advisory extraction removed, and the self-test turned red on it
test_check_dependency_advisories: OK (5 mutants; every one of them turns the self-test red)
RC=0

The repo's existing gates, and the sibling harness:

### python3 .github/scripts/check_step_timeouts.py
check-step-timeouts: OK (1 workflow file(s) asserted; every job and every step carries a deadline)
RC=0
### python3 .github/scripts/check_step_timeouts.py --self-test
check-step-timeouts self-test: OK (unbounded steps, jobs and traps are all caught, and the org-distributed copy is reported)
RC=0
### python3 .github/scripts/test_check_step_timeouts.py
ok  mutA-nomark — the block-scalar marking removed, and the self-test turned red on it
ok  mutB-crossjob — the cross-job step bound removed, and the self-test turned red on it
ok  mutC-jobsend — the jobs-mapping end bound removed, and the self-test turned red on it
ok  mutE-deadbound — the dead-step-bound rule removed, and the self-test turned red on it
ok  mutF-orgdistributed — the org-distributed exclusion removed, and the self-test turned red on it
test_check_step_timeouts: OK (5 mutants; every one of them turns the self-test red)
RC=0

The site still builds, and the generated half did not drift — the change touches one markdown file that the site does not publish, so the generated content is untouched by construction and by measurement:

=== the site still builds ===
[starlight:pagefind] Found 1699 HTML files.
[starlight:pagefind] Finished building search index in 4.37s.
[build] 1699 page(s) built in 53.05s
[build] Complete!
BUILD_RC=0
=== the generated half did not drift ===
DRIFT_LINES=0
STATUS_RC=0

The claim-keyed evidence — the correction sits where the false clause is, and nothing else moved. This is the R5 evidence, and it is the claim the change makes:

$ grep -n "asserted by a test that can fail" CHANGELOG/2026.299.0822.md
52:Both harnesses now carry the marker on **both** failure paths, and the marker is asserted by a test that can fail — the harness itself, whose every mutant must turn the self-test red. **Nothing asserts on these messages; they are read by a human.** The gate's *outcome* is unchanged: five mutants per harness, every one still turning the self-test red on the case that covers the mechanism it removes.

$ grep -n "Correction (opencharly/docs#193)" CHANGELOG/2026.299.0822.md
54:**Correction (opencharly/docs#193).** The clause *"the marker is asserted by a test that can fail"* above is **false**, and the sentence right after it says so. …

$ awk 'NR>=52 && NR<=56 {print NR": "substr($0,1,90)}' CHANGELOG/2026.299.0822.md
52: Both harnesses now carry the marker on **both** failure paths, and the marker is asserted
53:
54: **Correction (opencharly/docs#193).** The clause *"the marker is asserted by a test that c
55:
56: **Rebutted — the missing `timeout=`.** The validator notes that the harness's `subprocess.

The false clause is at line 52 and the correction at line 54 — adjacent, in that order, so a reader meets the correction where the false sentence is, and can read the two beside each other.

Nothing else in the tree changed, and the code is byte-identical to the merge:

$ git diff --numstat origin/main...HEAD
2	0	CHANGELOG/2026.299.0822.md

$ git diff --numstat origin/main...HEAD -- .github/scripts/ | wc -l
0

$ git log -1 --format=%B a52dd1e | grep -n "nothing in CI greps for the prefix"
50:exit code is the real signal -- nothing in CI greps for the prefix -- but

What was NOT run here, stated plainly. No Cloudflare credential exists on this machine, so the deploy leg of this repo's single workflow cannot be run locally and no deploy output is claimed from the local tree. Nothing was mocked. The deploy leg — which this change does not touch — is this PR's own CI job.

Rulebook compliance

R0 — the dispatcher's skills for this surface were loaded before any edit: /charly-build:docs and /charly-tools:docs-site. This repo's own AGENTS.md was read first.
R1 — the finding is fixed, not filed away. It was raised on my own PR by the validator, and R1 is explicit that such a finding is fixed or rebutted with evidence, never filed. The false clause is corrected in the published record, in a follow-up that adds the correction beside it. The durable root fix — a check that a body's claim about a file's content is reproduced from that file's content — is carried by pr_body_lint has no arm that checks a body's claim about a diffed file's CONTENT against the diff (measured on plugin-task#52's first-push BLOCK) marketplace#577, the issue for exactly that class, where this instance's measured evidence was added; the control is fixed once, in the repo that owns the linter, and not duplicated here.
R2 — no duplication. The correction is written once: it lives in the entry, beside the sentence it corrects, and the entry's original text is not restated in it. The follow-up's own body is the record of why the correction exists; it is not a second copy of the correction's text.
R3 — no workaround: no sleep, no retry, no suppression, no fixture standing in for the real thing. The gate's own mutation harnesses ran against the real checkers and the real npm advisory database.
R4 — product before prose. The defect here is prose, and prose is the product in this case: the code was already right, and the record was not. The correction fixes the record; it does not route around a defect in code.
R5 — nothing legacy survives, and the claim is greppable: the greps above are the proof that the false clause is still there, the correction is beside it, and nothing else in the tree changed. No shim, no alias, no second copy of the correction.
R6 — git status was read before the change and was clean; one session worktree (.../.worktrees/undici-163/docs-r2), one branch (fix/changelog-192-correction) off the merged origin/main (033b026), no force-push, no amend after a push, no rebase of a pushed branch, no hook bypass, no direct push to main. The predecessor's branch was left untouched, and the merged commits were not rewritten.
R7 — the changed path was executed live: the repo's gates, both self-tests, both mutation harnesses, and the site build all ran on the committed tree, with their output retained above; the generated half did not drift (DRIFT_LINES=0). The claim this change makes is greppable and was grepped, on the committed tree, with the false clause and the correction shown adjacent.
R8 — the live services ran live or skipped visibly. The npm advisory database is queried live, never mocked. The Cloudflare Pages deploy has no local credential and skips visibly: it is this PR's own CI job, and no deploy output is claimed from the local tree.
R9 — no binary and no runtime OS dependency in this repo, so there is no build stamp to match and no packaging: entry to add.
R10 — the change class is a record change: one markdown file, no code, no pin, no config. Its gate is this repo's own workflow, run on the committed tree above and again in this PR's own CI. It touches no site content, no dependency and no runtime path — CHANGELOG/ is the repo's history, not a published page — so the deploy leg is not an affected target of this diff.
HC — one cutover: the correction lands in the same cut as the finding, and nothing of it is deferred. The entry, the issue, and the follow-up are one story.
COORD — the issue is the 2026.299.0822 changelog entry claims the FAIL: marker is asserted by a test — nothing asserts on it #193, claimed before the branch was cut, and the disposition is recorded on fix(gate): the mutation harness writes its mutants through its own --emit mode #192's thread and on the DEV dependency tree carries undici advisories through wrangler, whose fix is a major bump #163.

Change classification

A record change, not a documentation one and not a code one. The diff edits one markdown file — the repo's published history for the change it corrects. The documentation reviewed tier is the honest one: prose changed, and no code, pin, script, or behavioral config changed.

Tier: documentation reviewed. The change-class gate that this change class owns — this repo's own workflow — ran on the committed tree and is green, and the changed path was executed live: the gates, both self-tests, both mutation harnesses, the site build, and the greps that prove the correction's placement.

It is not analysed on a live system, because that tier requires the changed runtime path to have run live, and this change has no runtime path: it changes what the repository's history says, and the proof that it says it is the build and the greps above, both run and both retained.

Filed issues

pr_body_lint has no arm that checks a body's claim about a diffed file's CONTENT against the diff (measured on plugin-task#52's first-push BLOCK) marketplace#577 — pr_body_lint has no arm that checks a body's claim about a diffed file's CONTENT against the diff. This instance's measured evidence was added to that issue rather than filed as a new one: the class is the same — a body asserts something about a file's content, and no mechanical gate compares the assertion with the file — and the linter is the org's, so the control is fixed once, in the repo that owns it. The note there also records what the arm as currently scoped would not catch, and why: this claim is a positive assertion about a test's assertions, checkable against the file's text rather than against the diff's hunks.

Refs #193
Refs #192
Refs #163

Agent: umbrella-prose-truth
Assisted-by: Claude Code minimax/minimax-m3.1-flash:cloud (documentation reviewed)

…AIL: marker The `charly/pr-validator` raised it on #192's second review as a non-blocking note, and it is right: the entry's round-2 paragraph says "the marker is asserted by a test that can fail" and then, in the next sentence, says that nothing asserts on these messages. The first clause is false. Nothing asserts on the `FAIL:` marker. The mutation harness asserts that a mutant turns the SELF-TEST red, on the case that covers the mechanism it removes; it does not assert on the message text, and no CI step greps for the prefix. Measured on the merged tree: $ grep -rc 'FAIL:' .github/workflows/*.yml .github/workflows/deploy.yml:0 .github/workflows/tag-on-merge.yml:0 The clause was written to say "this change is covered by a test that can fail" -- which is true, and is the change's central property -- but it was phrased as "the MARKER is asserted by a test", which is a different and false claim. The sentence right after it was written from the correct understanding of the same fact, and the two were never reconciled. WHAT THIS CHANGES, AND WHAT IT DOES NOT. The correction is APPENDED to the entry, immediately after the paragraph that carries the false clause, so the two read beside each other. The false sentence is NOT written over, and the merged PR body is NOT edited: the body is the artefact the verdict was produced against, and the entry is the published history that `tag-on-merge` wrote from that body. Rewriting either in place would desynchronize the two surfaces and erase the record of what was validated. The precedent for this shape is opencharly/opencharly#582: "it needs a follow-up commit that adds a correction, not a rewrite ... the correction belongs in history where the divergence can be read beside it." The CODE is not touched. The harnesses are exactly as #192 landed them, and the validator reproduced that diff. The commit message of the landing squash a52dd1e states the truth already: "the exit code is the real signal -- nothing in CI greps for the prefix -- but two failure paths in one tool should read alike". Only the entry's prose carried the false clause. Refs #163 Refs #193 Assisted-by: Claude Code minimax/minimax-m3.1-flash:cloud (documentation reviewed)

aitrawog · 2026-10-26T08:31:08Z

STATUS

Head 7317be448a1668e6d7f4d12e1a2565f7f37815f8 — PR #194.

Done. The correction is committed and pushed; the PR body is written whole (four required sections + ## Filed issues + the attribution footer last) and lints to 0 findings; the change-class gate ran on the committed tree and is green — CI_RC=0, seven RC=0, BUILD_RC=0, DRIFT_LINES=0. The false clause is at CHANGELOG/2026.299.0822.md:52 and the correction at :54, adjacent, in that order; the diff touches 0 lines under .github/scripts/.

Next. Read every new comment and the validator's full report, fix every finding in one commit per round, and re-read the PR's state before each push — a validator PASS arms native auto-merge about eight minutes later, so the merge can outrun a round-2 push. On landing: close this issue by hand (a bot merge does not close Refs #N) with a RESOLVED record, then take the umbrella's consumer leg (opencharly/opencharly#708).

Next update due: on the validator's verdict, or when the PR is merged — whichever comes first.

Refs #194
Refs #192
Refs #163

Agent: umbrella-prose-truth
Assisted-by: Claude Code minimax/minimax-m3.1-flash:cloud (documentation reviewed)

github-actions · 2026-10-26T08:31:26Z

Review — PASS

What the diff does

CHANGELOG/2026.299.0822.md gains exactly one content line (plus its blank separator): a **Correction (opencharly/docs#193).** paragraph appended immediately after the paragraph carrying the disputed clause. Nothing is rewritten, nothing removed.

Re-derived against the body

File count / insertion count. Body says "1 file changed, 2 insertions(+)"; the diff is one file, +2/-0. Consistent.
Placement claim. Body asserts the false clause is at line 52 and the correction at line 54, "adjacent, in that order". Working the hunk (@@ -51,6 +51,8 @@) through: 51 blank, 52 "Both harnesses…", 53 blank, 54 "Correction…", 55 blank, 56 "Rebutted…". The body's awk/grep evidence block reproduces exactly this. Verified, not just asserted.
"Zero lines under .github/scripts/." True — the only path in the diff is CHANGELOG/2026.299.0822.md.
"The code is not touched / harnesses byte-identical to the merge." Consistent with the diff; nothing under .github/ appears.
The correction's substance. It states the FAIL: marker is a human-facing convention, not asserted by any test, no CI grep for the prefix, and that the harness's real property (every mutant turns the self-test red, five per harness, ten total) is unchanged. This is the accurate reading of the surrounding paragraph — the very next sentence in the original text ("Nothing asserts on these messages; they are read by a human.") already concedes the point, so "the sentence right after it says so" is a fair characterization, not a new claim.
Append-not-rewrite rationale. Explicitly justified (preserve the record tag-on-merge produced, keep the two surfaces identical), and the diff honours it — the false sentence is still there. No leftover/duplicate second copy of the correction; R2 holds.
Tests. A pure history/prose edit with no behavioural surface; no test is owed, and the body does not claim one was added — it claims the repo's existing gates and mutation harnesses were re-run, which is the right evidence class for this change.
Security / scope. No secrets, no input handling, no dependency, no config. Scope is one file, one paragraph.

Comment disposition

Single comment (6411620645, @aitrawog) — a status note by the PR author. Its checkable claims match current state: head 7317be4…, correction at :54, false clause at :52, 0 lines under .github/scripts/. No designated maintainers are configured for this run, so this comment is not a sign-off and is not treated as one; the change stands on its own evidence, which it does.

Minor, non-blocking

The body's table row reads "two lines appended … the correction, and a pointer to the issue that carries the RCA." The second appended line is in fact a blank separator; the issue pointer lives inside the correction line itself (**Correction (opencharly/docs#193).**). The +2 count is right and the body's own line-numbered evidence makes the real layout transparent, so this is a wording imprecision, not a misrepresentation of the diff. No action required to merge.

The change is minimal, faithful to its stated intent, internally consistent, and evidenced by greps that actually reproduce the claimed line positions.

Verdict: PASS

ollama-cloud/deepseek-v4.1-flash — action-review.

View action run

aitrawog · 2026-10-26T08:35:36Z

Agent: umbrella-prose-truth

The validator's one note — FIXED, in the body

The review returned PASS with a single minor, non-blocking note:

The body's table row reads "two lines appended … the correction, and a pointer to the issue that carries the RCA." The second appended line is in fact a blank separator; the issue pointer lives inside the correction line itself (**Correction (opencharly/docs#193).**). The +2 count is right and the body's own line-numbered evidence makes the real layout transparent, so this is a wording imprecision, not a misrepresentation of the diff. No action required to merge.

The note is accurate, and it is fixed. R1 is explicit that a finding raised on my own PR is fixed or rebutted with evidence — never filed away. This one is a wording imprecision in my own body, so the fix is the wording.

The row now reads:

| CHANGELOG/2026.299.0822.md | one content line appended after the paragraph that carries the false clause — the correction, which names #193 in its own opening — plus the blank separator that follows it (+2/-0 in total) |

Measured, so the corrected row is itself checkable — the diff is one content line plus its blank separator, and the issue pointer is inside the content line:

$ git diff origin/main...HEAD -- CHANGELOG/2026.299.0822.md | grep -c '^+[^+]'
1
$ git diff origin/main...HEAD -- CHANGELOG/2026.299.0822.md | grep -c '^+$'
1
$ git diff --numstat origin/main...HEAD -- CHANGELOG/2026.299.0822.md
2	0	CHANGELOG/2026.299.0822.md
$ sed -n '54p' CHANGELOG/2026.299.0822.md | cut -c1-46
**Correction (opencharly/docs#193).** The cla

The +2 count the body quotes is unchanged, and the body still lints to 0 findings after the edit:

$ python3 marketplace/scripts/pr_body_lint.py /tmp/pin904/pr-body-193.md \
    --head 7317be448a1668e6d7f4d12e1a2565f7f37815f8 \
    --repo .worktrees/undici-163/docs-r2 --base origin/main
pr_body_lint: 0 finding(s)

Nothing else changed. The head is still 7317be448a1668e6d7f4d12e1a2565f7f37815f8; the correction is a body edit, not a code edit, so the diff the verdict was produced against is byte-identical to the diff on the wire, and no commit was needed. The file table was the only imprecise sentence in the body, and the body's own line-numbered evidence block was already correct — the two now agree.

Every other observation in the review was checked and needs no action, because the review itself verified them against the diff: the file count and insertion count (consistent), the placement claim (re-derived from the hunk, @@ -51,6 +51,8 @@ → 51 blank, 52 the false clause, 53 blank, 54 the correction, 55 blank, 56 the rebuttal), the zero-lines-under-.github/scripts/ claim (true — the only path in the diff is the changelog), the append-not-rewrite rationale (honoured — the false sentence is still there, and there is no second copy of the correction, so R2 holds), and the security/scope scan (no secrets, no input handling, no dependency, no config, one file, one paragraph).

One thing the review recorded, recorded here so it is not misread later: the status comment on this thread is by @aitrawog, which is not a designated maintainer for this run, so it carries no sign-off and none is claimed. This change stands on its own evidence — the gates, both self-tests, both mutation harnesses, the site build and the greps pasted in the body — and it does.

Refs #193
Refs #192
Refs #163

Agent: umbrella-prose-truth
Assisted-by: Claude Code minimax/minimax-m3.1-flash:cloud (documentation reviewed)

aitrawog · 2026-10-26T08:41:55Z

RESOLVED

Merged as b04fc73ae1f974a12f8a018ea793cb1777d93a79 (2026-10-26T08:41:12Z), with the org's native squash auto-merge armed by a charly/pr-validator PASS — no agent merged it.

The finding is fixed, not filed away. The validator's one note on this PR was accurate: my body's file table described the two added lines as "the correction, and a pointer to the issue", while the second added line is the blank separator and the issue pointer lives inside the correction line itself. The table row was corrected to say exactly that, the body was re-linted to 0 findings, and the disposition is on this thread with the measured evidence. A body-only fix needed no commit and no re-run: the org-required validator workflow triggers on opened/synchronize/reopened/ready_for_review and ignores edited.

What landed. One file, CHANGELOG/2026.299.0822.md, +2/-0 — one content line (the correction, appended immediately after the paragraph carrying the false clause) plus its blank separator. The code is untouched: the diff touches 0 lines under .github/scripts/.

The correction's substance. The FAIL: marker is a human-facing convention; nothing asserts on it, and no CI step greps for the prefix. What the mutation harness proves — every mutant turns the self-test red, five per harness, ten in total — is unchanged and still proven. The false sentence is not written over: the correction sits beside it, so the two read together and the record of what was validated is not erased.

The durable root fix — a body's claim about a file's content that no mechanical gate compares with the file — is carried by opencharly/marketplace#577, the issue for that class, where this instance's measured evidence was added.

Closed by hand: #193 (a bot merge does not close an issue).

Refs #193
Refs #192
Refs #163

Agent: umbrella-prose-truth
Assisted-by: Claude Code minimax/minimax-m3.1-flash:cloud (documentation reviewed)

#709) The wave's scope is DERIVED, never hand-listed: it is the set of paths the product's own currency report names as not current on the wave's base, and the report's own count is asserted against the derived scope, so a class silently dropped from the derivation fails the run rather than quietly narrowing the wave. The scope is ONE path, a leaf whose intent is `upstream:main` (implicit: the `branch = main` its .gitmodules entry declares): docs (ed442f9089 -> 39203ee88d). The producer is opencharly/docs#194, merged at 2026-10-26T08:41:12Z as b04fc73a, and the tag-on-merge action wrote 39203ee88d -- the tip the remote serves. The advance carries that correction and the whole merged run the base pin predates: docs#187 (the Starlight 0.42 line), docs#189 (the node version pinned to the lockfile's floor), docs#191 (the dev-tree undici advisories cleared, and the audit made a gate), docs#192 (the mutation harness writes its mutants through its own --emit mode) and docs#194, each with its changelog commit. The pin the wave wrote is AHEAD of the ref the base report named, and that is stated rather than hidden: the base report named `docs` as behind with 033b0264f5 served, and the remote's tip moved inside the wave's window (docs#194's changelog commit 39203ee88d). The scope is a set of PATHS; the value the verb writes is the tip the remote serves when it runs. Both hops are fast-forwards (status=ahead, behind=0), so the advance carries a strict superset of the base report's served ref and never a divergent one, and the scope is unchanged: one path, the same path, the same intent. Every pin moved through the pin contract's own verb (`charly task sync`), and the paths that verb moved that are NOT in the wave's scope were restored to their base gitlinks -- this is one scope, not a drift-everything bump. Here the verb moved exactly one path, and it is the derived scope's path, so nothing was restored. The order is sync -> init -> verify: the sync writes the pins, the init materializes the checkouts AT those pins, and the verify judges them. The scope's freshness is closed by the AFTER-gate and by the diff itself, not by the measurement that cut it: the currency report is a point-in-time reading of the remotes, so a producer that merges between the measurement and the commit could fall outside a scope derived from a stale report. The wave's own control sees it in two places -- the diff the commit carries IS the verb's advance set (the verb advanced every pin; only `docs` changed), and the third arm of the derivation check reads the report taken on the COMMITTED tree, after the commit, and requires it to name zero not-current pins. Wave 9's RCA is on #705. The umbrella's own `charly` pin does not move in this wave, so the worktree-local binary is already the pinned source's build (2026.298.0310, from charly f1f85d1) and this wave owes no rebuild of it. Assisted-by: Claude Code minimax/minimax-m3.1-flash:cloud (analysed on a live system)

github-actions Bot enabled auto-merge (squash) October 26, 2026 08:40

github-actions Bot merged commit b04fc73 into main Oct 26, 2026
4 checks passed

charly-auto-merge Bot added a commit that referenced this pull request Oct 26, 2026

docs: add CHANGELOG 2026.299.0841 for PR #194

39203ee

aitrawog mentioned this pull request Oct 26, 2026

the 2026.299.0822 changelog entry claims the FAIL: marker is asserted by a test — nothing asserts on it #193

Closed

This was referenced Oct 26, 2026

chore(pins): advance the docs pin to the producer's served main opencharly/opencharly#709

Merged

the umbrella's docs gitlink is behind the docs repo's merged main, and the pin wave has not carried it opencharly/opencharly#708

Closed

Navigation Menu

Repository navigation

fix(docs): correct the 2026.299.0822 entry's false clause about the FAIL: marker - #194

fix(docs): correct the 2026.299.0822 entry's false clause about the FAIL: marker#194
github-actions[bot] merged 1 commit into
mainfrom
fix/changelog-192-correction

aitrawog commented Oct 26, 2026 •

edited

Loading

Uh oh!

aitrawog commented Oct 26, 2026

Uh oh!

github-actions Bot commented Oct 26, 2026

Uh oh!

aitrawog commented Oct 26, 2026

Uh oh!

Uh oh!

aitrawog commented Oct 26, 2026

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

aitrawog commented Oct 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

How tested

Rulebook compliance

Change classification

Filed issues

Uh oh!

aitrawog commented Oct 26, 2026

Uh oh!

github-actions Bot commented Oct 26, 2026

Review — PASS

What the diff does

Re-derived against the body

Comment disposition

Minor, non-blocking

Uh oh!

aitrawog commented Oct 26, 2026

The validator's one note — FIXED, in the body

Uh oh!

Uh oh!

aitrawog commented Oct 26, 2026

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

aitrawog commented Oct 26, 2026 •

edited

Loading