@conarium-ai/core 0.2.45 → 0.2.47
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.md +88 -5
- package/docs/claims/reviews/0.2.46.md +67 -0
- package/docs/claims/reviews/0.2.47.md +44 -0
- package/docs/claims/surfaces.json +5 -0
- package/package.json +3 -1
package/CHANGELOG.md
CHANGED
|
@@ -1,5 +1,73 @@
|
|
|
1
1
|
# Changelog
|
|
2
2
|
|
|
3
|
+
## 0.2.47 - a number for the paper and the kit
|
|
4
|
+
|
|
5
|
+
v0.2.46 is on npm. Two things that are not that package now exist in
|
|
6
|
+
this repository, and publishing either of them under 0.2.46 would put
|
|
7
|
+
a new changelog and a new claims review on a number already sold.
|
|
8
|
+
|
|
9
|
+
**The two-ledgers preprint** (*Two ledgers, one window*) passed gate R2
|
|
10
|
+
on `docs/paper-two-ledgers`. `paper/` is not in the npm tarball. This
|
|
11
|
+
release does not republish the JSONL or exit-code rows already recorded
|
|
12
|
+
in 0.2.44–0.2.46.
|
|
13
|
+
|
|
14
|
+
**The starter kit** is `examples/starter-kit`: a local Postgres, three
|
|
15
|
+
mediated queries, signed receipts, and a reconcile against the
|
|
16
|
+
database's own counters. It is also not in the tarball. It is not a
|
|
17
|
+
second implementation.
|
|
18
|
+
|
|
19
|
+
`server.json` moves with `package.json` because the MCP registry reads
|
|
20
|
+
it and refuses a number it already holds.
|
|
21
|
+
|
|
22
|
+
`version_claim` asked npm with `--silent`. A 404 for a number that is
|
|
23
|
+
not on the registry arrived as an empty error, and the check printed
|
|
24
|
+
SKIP (registry unreachable) instead of "not tagged — nothing to
|
|
25
|
+
contradict". That is the state of every unpublished bump. The view is
|
|
26
|
+
no longer silent; E404 is "not published".
|
|
27
|
+
|
|
28
|
+
## 0.2.46 - three numbers that named something they were not
|
|
29
|
+
|
|
30
|
+
Every item here is a figure this project printed, published, or wrote down,
|
|
31
|
+
which turned out to measure something other than what it said. Two were
|
|
32
|
+
declared on the SCITT list before they were fixed; this is that debt.
|
|
33
|
+
|
|
34
|
+
**`exit_contract.mjs` reported its own work wrong.** The success line said
|
|
35
|
+
`27 invocation(s)`. It runs 25 processes; 27 was the pass counter, 23 cases
|
|
36
|
+
plus four assertions over two of them. A count that names something it is not,
|
|
37
|
+
printed by the file added to stop codes naming events they are not, in the
|
|
38
|
+
release that was about exactly that. Both numbers are now derived from the run
|
|
39
|
+
through a single `invoke()` wrapper and both are printed:
|
|
40
|
+
`25 invocation(s), 27 assertion(s)`. Add a case and the label moves on its own.
|
|
41
|
+
|
|
42
|
+
**`pack_locale.mjs` carried a figure its own subject forbids.** Its comment
|
|
43
|
+
said the letter pass finds 1041 where the changelog said 1034 — two
|
|
44
|
+
hand-written declarations of one number, inside the check whose whole purpose
|
|
45
|
+
is catching hand-written declarations drifting apart. Neither was reproducible:
|
|
46
|
+
on the 0.2.42 packed set the letter pass is 971 across 26, the word pass 310
|
|
47
|
+
across 20, and their union 1034 across 29. So 1034 was the union and never the
|
|
48
|
+
letter pass. The comment now describes the mechanism and states no figure at
|
|
49
|
+
all; the live number lives in `docs/claims/locale-residue.json`, where a run
|
|
50
|
+
compares it.
|
|
51
|
+
|
|
52
|
+
**And the number underneath both of them was a measurement of one disk.** See
|
|
53
|
+
the 0.2.43 entry, corrected in place: `324 / 22` came from a working tree
|
|
54
|
+
holding two gitignored files under `docs/audit/` that are in no commit. It
|
|
55
|
+
reached a changelog, and from there a public mailing list, where nobody can
|
|
56
|
+
reproduce it.
|
|
57
|
+
|
|
58
|
+
**`test/pack_tracked.mjs`** is the mechanism for that class, and it is not
|
|
59
|
+
about Turkish. Every path npm would ship must be tracked by git or fall under
|
|
60
|
+
one declared generated prefix, `dist/`, which `npm run build` reproduces from
|
|
61
|
+
source. Anything else is a file that exists on the packer's machine and in no
|
|
62
|
+
release, which means the package built here is not the package CI builds, and
|
|
63
|
+
every measurement taken over it is a measurement of somebody's disk.
|
|
64
|
+
|
|
65
|
+
The same mechanism once shipped a private key: `npm publish` from a machine
|
|
66
|
+
that had run the demo included `examples/_keys/audit-ed25519.pem`, which is why
|
|
67
|
+
package.json carries the sentence about .gitignore not binding npm and why
|
|
68
|
+
`pack_artefakt.mjs` locks that one path. This locks the class instead of the
|
|
69
|
+
path. It fails closed: if git cannot answer, the check does not pass.
|
|
70
|
+
|
|
3
71
|
## 0.2.45 - the rest of an outside report, and a check over the descriptions
|
|
4
72
|
|
|
5
73
|
**Behaviour change in `conarium-verify` and `conarium-coverage`.** Read the two
|
|
@@ -118,11 +186,26 @@ plainly rather than implied by an untested green.
|
|
|
118
186
|
|
|
119
187
|
`test/pack_locale.mjs` has always ended with a line naming what it does *not*
|
|
120
188
|
fix: Turkish left under `docs/`, out of interface scope, printed so it stays
|
|
121
|
-
visible. The line
|
|
122
|
-
**1034 lines across 29 files
|
|
123
|
-
|
|
124
|
-
|
|
125
|
-
|
|
189
|
+
visible. The 0.2.43 note wrote that line down as `324 occurrences across 22
|
|
190
|
+
files`, and called **1034 lines across 29 files** the letter pass. Remeasured
|
|
191
|
+
on 22 August 2026 against both bases that could have produced those figures:
|
|
192
|
+
the published 0.2.42 tarball (508 extracted paths) and `npm pack --dry-run` at
|
|
193
|
+
`v0.2.42` (256 paths). The lists differ only by `dist/` build artefacts, which
|
|
194
|
+
`docsRemainder` already excludes (`bin/`, `scripts/` and `public/` too). On
|
|
195
|
+
that packed-doc set the 0.2.42 instrument (`wordProbe` only) is **310 lines
|
|
196
|
+
across 20 files**. `324 / 22` reproduces on neither base, and the reason
|
|
197
|
+
was found rather than left open: two files under `docs/audit/` that are
|
|
198
|
+
gitignored, present on the machine the figure was taken on, and in no commit.
|
|
199
|
+
Copy them into a clean v0.2.42 worktree and the same guard prints 324 across
|
|
200
|
+
22; remove them and it prints 310 across 20. The number was never a fact about
|
|
201
|
+
the package. `.gitignore` does not bind npm, so a tree with extra files packs
|
|
202
|
+
a different package, and 0.2.46 adds `test/pack_tracked.mjs` to refuse that.
|
|
203
|
+
The `308 / 19` handover note below is the same class with the cause still
|
|
204
|
+
unknown. The **union** of letter and word on that set is **1034 / 29**; the
|
|
205
|
+
letter pass alone is 971 / 26. 1034 is the union, not the letter pass. The
|
|
206
|
+
remainder was counted by `wordProbe` alone — the pass written for Turkish that
|
|
207
|
+
carries *no* Turkish letter, and therefore the wrong instrument for Turkish
|
|
208
|
+
prose. Both passes now run and the union is reported.
|
|
126
209
|
|
|
127
210
|
What the package was shipping under that number: `docs/superpowers/`,
|
|
128
211
|
`docs/plans/`, `docs/specs/` and `docs/audit/` — dated internal design and
|
|
@@ -0,0 +1,67 @@
|
|
|
1
|
+
# Claims review — 0.2.46
|
|
2
|
+
|
|
3
|
+
- base: v0.2.45
|
|
4
|
+
- head: 696c694
|
|
5
|
+
- surface: sha256:e9d01e4e6f99670242922cf7c7d960d85adb9ca786d238d8a5f4279bc1c3eb43
|
|
6
|
+
- verdict: pass
|
|
7
|
+
- reviewer: maintainer, after a figure this file's process approved reached a
|
|
8
|
+
public archive and turned out to describe one machine
|
|
9
|
+
|
|
10
|
+
The surface hash is unchanged from 0.2.45. No file in `docs/claims/surfaces.json`
|
|
11
|
+
was touched: this release changes two guards, adds a third, and corrects a
|
|
12
|
+
changelog. That is exactly why the substantive finding below is not a finding
|
|
13
|
+
against the diff, and recording it here anyway is the point.
|
|
14
|
+
|
|
15
|
+
## Findings
|
|
16
|
+
|
|
17
|
+
none, against the claim surfaces.
|
|
18
|
+
|
|
19
|
+
## The reason this record is not empty
|
|
20
|
+
|
|
21
|
+
**A number that no claim-surface check could see was published to a mailing
|
|
22
|
+
list, and the process that let it through is this one.**
|
|
23
|
+
|
|
24
|
+
`324 Turkish lines across 22 packed doc files` was written into the 0.2.43
|
|
25
|
+
changelog and, on 22 August, into a message on the SCITT list. It is not
|
|
26
|
+
reproducible. Run v0.2.42's own `pack_locale.mjs` in a clean worktree and it
|
|
27
|
+
prints **310 across 20**. Copy in two gitignored files under `docs/audit/` that
|
|
28
|
+
sit on one machine and in no commit, and it prints **324 across 22**. The
|
|
29
|
+
figure measured a working directory.
|
|
30
|
+
|
|
31
|
+
Every gate held and none of them was pointed here:
|
|
32
|
+
|
|
33
|
+
- `claim_discipline` scans surfaces for retracted phrasings. A number is not a
|
|
34
|
+
phrasing.
|
|
35
|
+
- `denetci` compares the surface hash. `CHANGELOG.md` is not a surface, on
|
|
36
|
+
purpose, because a changelog is a record rather than a claim about the
|
|
37
|
+
present.
|
|
38
|
+
- `pack_locale` printed the figure. It had no reason to doubt its own output,
|
|
39
|
+
and it was correct about the tree it was given.
|
|
40
|
+
- `version_claim` compares what npm shipped against the tag. It compares
|
|
41
|
+
contents, not the provenance of the files that produced them.
|
|
42
|
+
|
|
43
|
+
So the defect was not that a check failed. There was no check whose subject was
|
|
44
|
+
*"is the tree this measurement was taken over the tree this repository
|
|
45
|
+
describes"*. `test/pack_tracked.mjs` is that check, and it is the whole
|
|
46
|
+
deliverable of this release. The two label repairs are consequences.
|
|
47
|
+
|
|
48
|
+
**What it does not close.** It establishes that the packed set matches the
|
|
49
|
+
repository plus `dist/`. It says nothing about figures measured over anything
|
|
50
|
+
else — a corpus, a database, a run on another machine — and nothing about
|
|
51
|
+
whether a number in prose was copied correctly from a run that did happen. A
|
|
52
|
+
figure written by hand into a document is still a hand-written figure; the only
|
|
53
|
+
one now mechanically pinned is the locale remainder, in
|
|
54
|
+
`docs/claims/locale-residue.json`.
|
|
55
|
+
|
|
56
|
+
## Not a finding, recorded so the next reader does not re-derive it
|
|
57
|
+
|
|
58
|
+
The first attempt to show `pack_tracked` red used the two `docs/audit/` files
|
|
59
|
+
that caused the original defect. It stayed **green**, and correctly: 0.2.43
|
|
60
|
+
excluded `docs/audit` from the package, so those files are no longer packed
|
|
61
|
+
from any tree. The red was produced instead with a stray file under `docs/`,
|
|
62
|
+
which is where the class still bites.
|
|
63
|
+
|
|
64
|
+
Worth stating because the near-miss is instructive: a red test built from the
|
|
65
|
+
historical culprit can pass for a reason that has nothing to do with the check
|
|
66
|
+
being wrong, and a reviewer reading only the green would have concluded the
|
|
67
|
+
guard worked. It did work. The demonstration did not.
|
|
@@ -0,0 +1,44 @@
|
|
|
1
|
+
# Claims review — 0.2.47
|
|
2
|
+
|
|
3
|
+
- base: v0.2.46
|
|
4
|
+
- head: f2372c9
|
|
5
|
+
- surface: sha256:b3c49849a617eb4dfab1fe4427a40f7de440516d30bf62894506be2c9c826e08
|
|
6
|
+
- verdict: pass
|
|
7
|
+
- reviewer: Claude, after the release branch merged `docs/draft-07`. The first
|
|
8
|
+
version of this record was written at `ea09896` by Cursor, before the draft
|
|
9
|
+
landed, and the gate refused it: the surfaces it described were not the ones
|
|
10
|
+
about to ship. Re-read from `v0.2.46` over the full set, `standards/README.md`
|
|
11
|
+
and `standards/draft-dogru-scitt-disclosure-evidence-07.md` included.
|
|
12
|
+
|
|
13
|
+
The version number moved because a new changelog section and this record
|
|
14
|
+
would otherwise ship under a number already on npm. `server.json` moves
|
|
15
|
+
with `package.json`. The starter kit landed as `examples/starter-kit` and
|
|
16
|
+
is excluded from the tarball. The two-ledgers preprint remains on
|
|
17
|
+
`docs/paper-two-ledgers`; `paper/` is not in this tree and is not in the
|
|
18
|
+
tarball. `-07` was read against the four neighbouring drafts it cites
|
|
19
|
+
(datatracker, 23 August 2026), against the -04 rewrite published by the
|
|
20
|
+
reviewer it thanks (no sentence of his entered the text), and its sentence
|
|
21
|
+
counts were recomputed with `test/draft_sentences.mjs`.
|
|
22
|
+
|
|
23
|
+
## Findings
|
|
24
|
+
|
|
25
|
+
### 1. standards/ADOPTION-EVIDENCE.md:20 — Vaara "adopted the wording"
|
|
26
|
+
|
|
27
|
+
- Claim: "Adopted the wording 'removal is detectable relative to a retained head' on the Vaara conformance / evidence page after the 21 August 2026 list exchange." — kind `wording adopted`
|
|
28
|
+
- Mechanism: the SCITT list archive and the Vaara page would show the sentence originating here and appearing there.
|
|
29
|
+
- Why it falls short: the sentence was proposed by Iman Schrock on the list (20 August 2026, `arch/msg/scitt/I5mdtYRinS-wjrnbwCnyXwwvrjU`) and endorsed in this project's reply the next morning. It appears nowhere in this repository before 23 August 2026. Vaara's `reproductions.json` credits this project with the observation that a truncated chain verifies clean, not with the wording. A citation was labelled an adoption.
|
|
30
|
+
- Disposition: reworded in `f2372c9` — kind `citation`, with the origin of the sentence stated and the three source URLs given.
|
|
31
|
+
|
|
32
|
+
### 2. standards/ADOPTION-EVIDENCE.md:19 — Pablo Play "merged a partition"
|
|
33
|
+
|
|
34
|
+
- Claim: "Merged a partition of the receipt-set completeness discussion on the SCITT list (thread 'Closing omission', 21 August 2026)." — kind `citation`, URL "not retrieved in this pass"
|
|
35
|
+
- Mechanism: a list message from that sender citing this draft.
|
|
36
|
+
- Why it falls short: his three list messages of 19–20 August 2026 concern argentum-core and the coverage-attestation draft; none cites this draft. The row had no URL because there was nothing to point at.
|
|
37
|
+
- Disposition: row removed in `f2372c9`, with a note saying so. A row for Joel Hillier's draft announcement — whose own words call his disposition vocabulary this draft's standing ladder "now aimed at a very different question" — replaces it, with its archive URL.
|
|
38
|
+
|
|
39
|
+
### 3. standards/ADOPTION-EVIDENCE.md:21 — Elara row without a reachable URL
|
|
40
|
+
|
|
41
|
+
- Claim: "Recorded this project's OpenSSL / ML-DSA-65 observation in Elara `SIGNING.md`" — URL "not re-fetched in this pass".
|
|
42
|
+
- Mechanism: the file itself.
|
|
43
|
+
- Why it falls short: it is a claim about a document nobody in this pass had opened. The file exists at `site/receipts/SIGNING.md` and names Emek Can Doğru and Conarium as the first verification of Elara bundles with non-Elara code — which is this project verifying Elara, the reverse direction from an adoption.
|
|
44
|
+
- Disposition: URL filled and the direction stated in `f2372c9`. `standards/ADOPTION-EVIDENCE.md` added to `docs/claims/surfaces.json` so the next row is read by the gate.
|
|
@@ -110,6 +110,11 @@
|
|
|
110
110
|
"path": "docs/security/NPM-PROVENANCE.md",
|
|
111
111
|
"why": "Ships in the tarball and tells a reader how to check where the package came from. Unlisted until 0.2.33, by which point it had gone stale in the worst direction: it said 'this is not a published release' inside eight published releases, and pointed at a verification command that answers 404 for npm provenance. A page about how to distrust us is one a reader must be able to trust.",
|
|
112
112
|
"phrasing": true
|
|
113
|
+
},
|
|
114
|
+
{
|
|
115
|
+
"path": "standards/ADOPTION-EVIDENCE.md",
|
|
116
|
+
"why": "Public on GitHub and linked from standards/README.md. Every row asserts what an outside party did with this draft; a row that overstates a citation as an adoption is the same overclaim the kind labels exist to prevent. Its first version carried one.",
|
|
117
|
+
"phrasing": true
|
|
113
118
|
}
|
|
114
119
|
]
|
|
115
120
|
}
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@conarium-ai/core",
|
|
3
|
-
"version": "0.2.
|
|
3
|
+
"version": "0.2.47",
|
|
4
4
|
"mcpName": "io.github.dogrucanemek-alt/conarium",
|
|
5
5
|
"//files-artefakt": "Generated files under examples/ (private key, receipt/audit JSONL) are EXCLUDED. .gitignore does not bind npm: because the allow-list contained 'examples', `npm publish` on a machine that had run the demo was shipping _keys/audit-ed25519.pem (PRIVATE KEY) and the receipts. test/pack_artefakt.mjs locks this. The exclusions are LIMITED to examples/: receipts.jsonl and .pub.pem under test-vectors/ MUST ship, or a third party cannot run the conformance vectors.",
|
|
6
6
|
"//files": "Allow-list of what enters the package. Without it npm sent everything: 296 files / 1.5 MB — CI workflows, gitleaks config, a 137 KB marketing image, the site's api/chat.js. A bank security team opens the tarball and sees those. scripts/ is here ON PURPOSE: conarium-reconcile references scripts/pg-snapshot.sql at runtime; without it the user files a support mail.",
|
|
@@ -28,6 +28,8 @@
|
|
|
28
28
|
"!examples/**/*.jsonl",
|
|
29
29
|
"!examples/**/*.keyid",
|
|
30
30
|
"!examples/**/conarium.tokens.json",
|
|
31
|
+
"!examples/starter-kit",
|
|
32
|
+
"!examples/starter-kit/**",
|
|
31
33
|
"conformance",
|
|
32
34
|
"test-vectors",
|
|
33
35
|
"!test-vectors/jcs/rfc8785",
|