Repository navigation
feat(registry)!: re-mint ten placeholder UUIDs as UUIDv8 and re-register (ADR 0002) - #107
Merged
Merged
Conversation
…ter (ADR 0002) Ten packages were registered under hand-typed placeholder UUIDs. Remove each old binding and the versions under it, and register the package again under a fresh UUIDv8 (ADR-008 profile T) at a patch bump: HardwareResilience, SiliconCore, ZeroProb, QuantumCircuit, ShellIntegration, FirmwareAudit, ViperPkg and LowLevel at 0.1.1; Causals and SMTLib at 0.2.1. Remove InvestigativeJournalist 0.1.0, whose Deps bind the retired Causals and ZeroProb UUIDs. Register InvestigativeJournalism 0.1.1, the renamed upstream, as a new package. Every new entry is generated with RegistryTools v2.4.3 from the Project.toml at the registered commit. SMTLib's nine placeholder weak dependencies are left out of its WeakDeps.toml. ViperPkg's repo URL moves to hyperpolymath/jtv-lang. BREAKING CHANGE: a Manifest that pins one of the eleven retired UUIDs no longer resolves against this registry. Run Pkg.rm and then Pkg.add of the same name; InvestigativeJournalist users add InvestigativeJournalism. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016XJoh8YA2sCWQmq93CF8RX
Contributor
|
Warning Review limit reachedYou've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. Next included review available in 52 minutes. View limit detailsLimit details: You’ve used the included review currently available. Review configuration: ⚙️ Run configuration
📒 Files selected for processing (57)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Each new version registers the default-branch tree, not the old tree under a new UUID, so it carries every change merged since the previously registered version. Record the measured drift: 88-219 files per package, mostly CI and docs scaffolding; Julia source changes small; no exported name removed, three added in Causals; non-Julia src/abi scaffold dropped in four packages. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016XJoh8YA2sCWQmq93CF8RX
This was referenced Oct 9, 2026
hyperpolymath
added a commit
that referenced
this pull request
Oct 9, 2026
…21/D322) (#109) <!-- SPDX-License-Identifier: CC-BY-SA-4.0 --> ## Summary This PR registers the nine Marid Tier-1 packages at **v0.1.0** (owner rulings D321/D322). Their source is the `metadatastician/marid` monorepo, and each is registered by `subdir = "packages/<Pkg>"` at commit `fd139f763a306a7e2eca60d96ea1a1a9f77c1e71`. That commit is the UUIDv8 profile-T re-mint, merged as metadatastician/marid#78. Packages: MaridArango, MaridCodec, MaridControl, MaridCRDT, MaridIR, MaridLive, MaridRaft, MaridStorage, MaridTransport. The PR also widens `ci/validate.jl` Layer 2b. Marid lives under the `metadatastician` organisation, and the unpatched validator rejects all nine packages with an owner mismatch. This registry is the only target. Nothing here is aimed at JuliaRegistries/General. ## Merge order: this PR is a draft until both conditions hold 1. **The nine `<Pkg>-v0.1.0` tags exist at `fd139f7`.** They are owner-signed. The `MaridIR-v0.1.0` canary goes first, because `release.yml` has never run before. 2. **#107 (UUIDv8 re-mint of ten placeholder UUIDs) has merged.** Both PRs edit `Registry.toml` and `CHANGELOG.adoc`. A trial merge (`git merge-tree` of #107's head `cf67a95e` with this branch) conflicts in both files. Once #107 lands, I will: - merge `main` into this branch (no rebase, so the signed commit survives); - resolve the two files; - re-run the validator; - update the head SHA below. ## Type of change - [x] Registry change (new packages) - [x] CI change (validator owner allow-list) - [x] Docs (README package table, CHANGELOG, REGISTRY-POLICY) ## 📌 New pins - **PR head SHA: `c7f9f64b6aec27412a8d7f76488480f57a3b7cad`** - **Upstream commit: `fd139f763a306a7e2eca60d96ea1a1a9f77c1e71`** (metadatastician/marid `main`, #78) - `git-tree-sha1` per package. Each one equals `git rev-parse fd139f7:packages/<Pkg>`, checked one by one: | Package | UUID (v8) | git-tree-sha1 | |---|---|---| | MaridArango | `01a115ae-6523-8273-8497-753822252d6a` | **`5c71ca32eb207776d2dda2969570e8c640bfae14`** | | MaridCodec | `01a115b0-6977-8194-b1bb-ba277a2ddf75` | **`122522a80a17e8165d3e0230bcb3196b4b32c7e9`** | | MaridControl | `01a115b1-6b89-8bc1-b693-306d1d0301d5` | **`0ef98fceb85a30faa4e7e64a88c23094dfa68948`** | | MaridCRDT | `01a115af-6777-83e7-bbf4-4312fe84fb45` | **`432bf84d4b482e84b5a9f0afa6e271e522bdd252`** | | MaridIR | `01a115b4-723a-8911-b469-7c18b764c41a` | **`4c76daa8e4ac08d02db18688557e2341b867e1e4`** | | MaridLive | `01a115b5-7479-8d82-a0d2-d2a681675e27` | **`367e0e315aa1ab0b767fae6dadbecbd25a21e6a9`** | | MaridRaft | `01a115b8-7a8a-8f24-baaa-cf06974b0145` | **`e1e8405aeb1cb5742ee346dbc74c56c7e936baef`** | | MaridStorage | `01a115b9-7cf4-8463-9dfb-224e0c6ab613` | **`1177666f911e55c25bfdf3058ea5a49144f8c5d3`** | | MaridTransport | `01a115ba-7f33-8545-9989-732142b76ec2` | **`e7ba55688bd13c422628f70f2ba921a4e769d21b`** | The PR adds or changes no action `uses:` lines and no `actions.lock` entries, lockfiles or container digests. ## Changes - `M/<Pkg>/{Package,Versions,Compat,Deps}.toml` for the nine packages. These are LocalRegistry 0.5.7 output, unchanged. It was run in a throwaway clone because LocalRegistry refuses a worktree. - `Registry.toml`: **+9 lines only**, inserted by hand between `LowLevel` and `PRComms`. - **Disclosure:** LocalRegistry rewrites the whole `[packages]` table sorted by UUID. I discarded that rewrite so the diff shows only the nine new entries. - The next LocalRegistry run on `main` will therefore produce the re-sort as unrelated churn. That churn is expected and is not a defect. - `ci/validate.jl` Layer 2b: - Adds `ALLOWED_PACKAGE_OWNERS = ("hyperpolymath", "metadatastician")` and a documented `allowed_package_owner(slug)`. A package's `Package.toml` `repo` may now be owned by either. - The registry's own `repo` must still be `hyperpolymath/…`. - Also fixes a misplaced docstring: the `check_repo_slug!` docstring sat on `extract_github_slug`. Each function now has its own, confirmed with `Docs.doc`. - `docs/REGISTRY-POLICY.adoc`: - The Layer 2b paragraph now names both owners. - The stale "across all 37 packages currently registered" now reads "every package registered in `Registry.toml`". The base had 32 packages, and this PR makes 41. - `README.adoc` package table and `CHANGELOG.adoc` `[Unreleased]` updated. ## RSR Quality Checklist ### Required - [x] No banned language patterns. The PR adds only TOML, Julia and AsciiDoc; no JS, TS, Python or Go. - [x] SPDX headers on new or modified source files. - The modified files (`ci/validate.jl`, both `.adoc` files) keep their headers. - The new `M/Marid*/*.toml` files carry none, which matches 29 of the 32 existing package dirs. Registry TOML is generated data, so I did not add headers by hand. - [x] No secrets, credentials or `.env` files. - [ ] CI is green. Not yet run: CI starts when this PR opens. I will update this box from the real rollup. ### As Applicable — Registry Changes - [x] `Registry.toml` index updated (+9). - [x] Per-package TOMLs are consistent. Layer 2a found 0 name↔UUID errors, and every tree hash matches upstream (see above). - [x] Upstream meets the GOVERNANCE quality bar. - Each `packages/<Pkg>` has its own `LICENSE`, which marid's `release.yml` checks per subdir. - `[compat]` is upper-bounded for every dep and weakdep (marid#22/D68 gate). - The UUIDs are the v8 profile-T re-mint. - [ ] Upstream tag exists and matches. **Not yet.** The tags `<Pkg>-v0.1.0` at `fd139f7` are owner-signed (EstateTagging `creation` rule; GITHUB_TOKEN cannot create tags). They are to be pushed **before this PR merges**, and I will tick this box once `git ls-remote --tags` shows all nine. ### As Applicable — Documentation / Infra Changes - [ ] `.machine_readable/6a2/*.a2ml`: n/a. A2ML is retired estate-wide (D308), so these boxes are not applicable and no `.a2ml` was created. - [ ] `TOPOLOGY.md`: n/a, the architecture is unchanged. - [x] CHANGELOG updated. This repo's changelog is `CHANGELOG.adoc`; the template's `.md` name is stale. - [ ] Wiki: not updated in this PR. The Packages wiki page needs the nine rows added; that is an owner wiki edit. ## Testing **1. Validator, Julia 1.10.10 (the CI version), this branch:** ``` julia --project=<env: LocalRegistry 0.5.7, RegistryCI 11.0.0> ci/validate.jl <worktree> rc=0 Layer 1 (RegistryCI, informational): 670 passed, 22 failed Layer 2b: Checked repo field on 41 package Package.toml files (0 owner mismatches) Custom-check errors: 0 — VALIDATION PASSED (custom checks) ``` **2. Same validator, base `9e4dc83`:** rc=0, Layer 1 530 passed / 20 failed, 32 packages, 0 mismatches. - Every Layer 1 failure on both sides is `depuuids ⊆ alluuids`. That is the known gap for packages that depend on General. - This PR adds exactly **2** of them: MaridCodec depends on JSON3, and MaridTransport depends on HTTP. Layer 1 does not gate the exit code. **3. Controls proving Layer 2b is live (not vacuous):** - **Mutant:** MaridIR `repo` set to `evilcorp/marid` gives rc=1 with `MaridIR: Package.toml repo resolves to slug "evilcorp/marid", expected an owner in ("hyperpolymath", "metadatastician")`. - **Unpatched validator** (base `ci/validate.jl`) on this tree gives rc=1 with 9 owner mismatches. So the validator change is necessary, not cosmetic. **4. End-to-end install** (fresh depot, Julia 1.10.10, clone of this branch at `c7f9f64`): ```julia using Pkg Pkg.Registry.add("General") Pkg.Registry.add(Pkg.RegistrySpec(url = "<clone of feat/marid-tier1>")) P = ["MaridIR", "MaridCodec", "MaridControl", "MaridCRDT", "MaridLive", "MaridRaft", "MaridStorage", "MaridTransport", "MaridArango"] Pkg.add(P) for p in P; @eval using $(Symbol(p)); println("LOADED ", p); end ``` Result: rc=0. **All nine** resolve at v0.1.0 (`[01a115ae]` … `[01a115ba]`), precompile, and load (9 × `LOADED`). The General deps come from General: JSON3 for MaridCodec, and HTTP v1.11.0 for MaridTransport. --- 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01WPSJ7fBhVAMcpSffCBWUDo https://claude.ai/code/session_013omQK26s4uDjJMkdqNEvEG Signed-off-by: Jonathan D.A. Jewell <6759885+hyperpolymath@users.noreply.github.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
5 of 12 tasks
Registry.toml: keep this branch's re-minted LowLevel binding (01a11d29-d884-…) and main's nine Marid entries. CHANGELOG.adoc [Unreleased] Added: keep both sides' entries. Validated on the merged tree: ci/validate.jl rc=0, 41 packages, custom checks 0 errors / 0 warnings; Layer 1's failure set equals main 7c52894's apart from InvestigativeJournalist -> InvestigativeJournalism. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016XJoh8YA2sCWQmq93CF8RX
|
hyperpolymath
added a commit
that referenced
this pull request
Oct 9, 2026
…003 (#110) <!-- SPDX-License-Identifier: CC-BY-SA-4.0 --> ## Summary #109 registered the nine Marid Tier-1 packages and, in the same change, widened `ci/validate.jl` Layer 2b so a package `repo` may be owned by `hyperpolymath` **or** `metadatastician`. It carried no ADR. This PR adds that ADR. It records the decision #109 implemented and makes one rule explicit: the set of allowed owners is **closed**, so adding an organisation takes another ADR. It contains **no code change**. `ci/validate.jl` differs from `main` only in the comment on `ALLOWED_PACKAGE_OWNERS`. On 2026-10-09 the owner ruled that the metadatastician widening be recorded in its own PR with ADR 0003. This PR was opened before #109 merged, with its own implementation. When #109 landed first, I merged `main` in, kept main's code (`EXPECTED_SLUG_OWNER`, `ALLOWED_PACKAGE_OWNERS`, `allowed_package_owner`) and dropped my parallel constants. What remains is the record. It changes registry policy text, so it needs BDFL review. **Automerge is not armed.** > [!NOTE] > **Merge order:** none. `git merge-tree` of this head against #107's head `b1563013` is clean (rc=0), so the two land in either order. ADR numbers do not collide: #107 carries ADR 0002 and this PR carries 0003. ## Changes - **`docs/decisions/0003-package-repos-in-metadatastician.adoc`** (new): - context: D54-A, plus D321 (stage 1 of D67-A on marid#21) and D322; - three decisions: 1. packages may be owned by either organisation; 2. the set is closed; 3. Registry.toml's own `repo` stays `hyperpolymath`; - three alternatives rejected; - the consequences. The ADR names #109 / `7c52894` as the implementation. - **`docs/REGISTRY-POLICY.adoc`**: one sentence after the Layer 2b paragraph says the set is closed and links ADR 0003. - **`ci/validate.jl`**: the comment on `ALLOWED_PACKAGE_OWNERS` gains the same sentence. No executable line changes. - **`CHANGELOG.adoc`** [Unreleased] → Added: one entry for the ADR, directly below #109's Marid entry. ## 📌 New pins - **Head SHA:** **`8d7eeb362b4b676b27eae1abfffea401af97bf2d`** (a signed merge of `main` `7c52894`; this branch's other commit, `a4a4bef`, is signed too). - **Action pins:** none. No workflow, `actions.lock`, lockfile or container digest changes. ## RSR Quality Checklist ### Required - [x] No banned language patterns: AsciiDoc, plus one comment in an existing Julia file. - [x] SPDX license headers on new files: the ADR has `CC-BY-SA-4.0`, matching ADR 0001. - [x] No secrets, credentials or `.env` files. - [ ] CI is green: **not entirely**. Three non-required checks are red on `main` at `7c52894` too (and at `9e4dc83` before it), so this PR does not cause them. They are deferred below. ### Deferred checks (§5c) - `actions.lock is in sync with the workflow YAML`: deferred to #108. Dependabot #106 bumped `smtp-notify-action` to v0.5.0 without refreshing `.github/workflows/actions.lock`. - `governance / Allowlist Preflight`: deferred to #105. `GH_TOKEN` is empty in the step, so it reports `could not read live Actions permissions`. - `governance / Workflow security linter`: deferred to #105. `scripts/update-actions-lock.sh: No such file or directory`, exit 127. ### As Applicable — Registry Changes - [ ] `Registry.toml` index updated: N/A. No package is added, removed or rebound. - [x] Per-package TOMLs consistent: the validator reports 0 custom errors on this branch (Testing below). - [ ] Upstream package meets the quality bar: N/A. No package is registered here. - [ ] Upstream tag exists and matches: N/A. No version is registered here. ### As Applicable — Documentation / Infra Changes - [ ] `.machine_readable/6a2/*.a2ml`: N/A. A2ML is retired estate-wide (D99/D269c), so no `.a2ml` file is created or edited. - [ ] `TOPOLOGY.md`: N/A. No architectural change. - [x] Changelog updated. This repo's file is `CHANGELOG.adoc`. - [ ] Wiki updated: N/A. The wiki does not describe Layer 2b. ## Testing Every run is `ci/validate.jl <tree>` on Julia 1.10.10 with RegistryCI v11.0.0, against a `git ls-files` copy of this head. Each copy is run with `GIT_CEILING_DIRECTORIES` set, so no enclosing repository's `origin` leaks into the slug comparison. | Tree | Result | |---|---| | This head (41 packages, nine `metadatastician/marid`) | **rc=0**. Custom checks 0 errors / 0 warnings | | Same, MaridIR `repo` → `evilcorp/marid` (planted negative) | **rc=1**: `MaridIR: Package.toml repo resolves to slug "evilcorp/marid", expected an owner in ("hyperpolymath", "metadatastician")` | | ADR claim "every package registered before #109 is `hyperpolymath`-owned" | `repo` owners of the 32 non-Marid `Package.toml` files on `main`: **32 × `hyperpolymath`** | | ADR claim "the old check reported one mismatch per Marid package" | #109's tree with `main`'s pre-#109 validator (`9e4dc83`): **rc=1, 9 owner mismatches** | The only executable file touched is `ci/validate.jl`, and only in a comment. `standards/.githooks/docstring-scan.sh` skips Julia files, so it gives no verdict here, and no function is added or changed. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_016XJoh8YA2sCWQmq93CF8RX Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Ten packages in this registry were registered under hand-typed placeholder UUIDs, runs of consecutive hex such as
d5f6a7b8-c9d0-4e1f-9a2b-3c4d5e6f7a8b. This PR removes each old binding and the versions under it, then registers the package again under a fresh UUIDv8 (ADR-008 profile T) at a patch bump.It also removes InvestigativeJournalist 0.1.0 and registers its renamed upstream, InvestigativeJournalism 0.1.1, as a new package. The old entry's
Deps.tomlbinds the retired Causals and ZeroProb UUIDs.This is a UUID/name binding change, so per
docs/REGISTRY-POLICY.adocit carries an ADR (docs/decisions/0002-uuidv8-remint-and-reregistration.adoc), needs BDFL review, and must not auto-merge. Automerge is not armed.The owner's rulings:
Note
Merged 2026-10-09T00:24:48Z as
e8383fb(signed byweb-flow). Its tree,fbbd8f01, is the tree of headb1563013, the head this body describes. This body was updated after the merge; at merge time it still described the earlier headcf67a95.The merge order held. Both upstream PRs merged first, and each squash commit's root tree is exactly the tree registered here:
hyperpolymath/SMTLib.jl#48merged at 00:23:53Z as487c7ab4, tree05a805f3…;hyperpolymath/InvestigativeJournalism.jl#78merged at 00:24:24Z as2cc30882, tree6e624d20….The other nine trees were already on their default branches.
Checked after merge on
main991a5b7(this PR plus #110):ci/validate.jlgives rc=0: 41 packages, 0 custom errors.Pkg.addof all 11 packages into a fresh depot that holds only General andmain: 11/11, each at its registered UUID, version and tree, fetched from upstream.#109 (Marid Tier-1) merged before this PR.
main7c52894was merged into this branch atb1563013, keeping both the re-minted LowLevel binding and the nine Marid entries. No MaridDeps.tomlnames a retired UUID.Changes
0.1.0);0.2.0);0.1.0).0.1.1;0.2.1.0.1.1(01a11de2-b9d3-80b4-b616-640df22918c0) atI/InvestigativeJournalism.repo→https://github.com/hyperpolymath/jtv-lang.git(subdirviper-pkg). The oldjulia-the-vipername redirects there.update_package_file,update_versions_file,check_deps!/update_deps_file,check_compat!/update_compat_file), fed theProject.tomlat the registered commit. SMTLib declares nine placeholder weak deps (11111111-…to99999999-…) that name no package; they are left out of itsWeakDeps.toml. No other input was filtered.CHANGELOG.adoc[Unreleased]: Added / Changed (marked breaking) / Removed;README.adoc: the 11 version cells.📌 New pins
b15630137b1d915d42fd1b4013c33089d959e73c. It is a signed merge ofmain7c52894intocf67a95; this branch's own commits,3eff90fandcf67a95, are signed too.uses:,actions.lockor container digest changes; no workflow file is touched.6e624d2077ca172390591328e5faaf894c2b0d21(InvestigativeJournalism.jl#78). SMTLib05a805f35b2b6f235e0d0585c58d594a3f2b219eis now on SMTLib.jlmainas squash487c7ab4.RSR Quality Checklist
Required
CC-BY-SA-4.0.main, 1 of the 47 touched package TOMLs carries a header, and here 1 of 51 does. The one existing header is kept and no new ones are invented, matching Registrator's output..envfiles.scan / gitleaks, passes, and so doesvalidate(this repo's registry check) along with 22 other checks. Three non-required checks are red, and each is equally red onmainat9e4dc83and at7c52894, so this PR did not cause them. They are deferred below.Codeac analyze resultsis a legacy status that has not reported.Deferred checks (§5c)
actions.lock is in sync with the workflow YAML: deferred to CI: actions.lock out of sync after Dependabot #106 bumped smtp-notify-action to v0.5.0 #108. Dependabot chore(deps): bump hyperpolymath/smtp-notify-action from 0.3.0 to 0.5.0 in the actions group #106 bumpedsmtp-notify-actionto v0.5.0 without refreshing.github/workflows/actions.lock.governance / Allowlist Preflight: deferred to CI: governance linter calls a missing scripts/update-actions-lock.sh; Allowlist Preflight runs with an empty GH_TOKEN #105.GH_TOKENis empty in the step, so it reportscould not read live Actions permissions.governance / Workflow security linter: deferred to CI: governance linter calls a missing scripts/update-actions-lock.sh; Allowlist Preflight runs with an empty GH_TOKEN #105.scripts/update-actions-lock.sh: No such file or directory, exit 127.As Applicable — Registry Changes
Registry.tomlindex updated: 11 entries rebound, IJ-ist→IJ-ism. With feat(registry): register the nine Marid Tier-1 packages at v0.1.0 (D321/D322) #109's nine Marid entries frommain, 41 entries, as onmain.ci/validate.jlcustom layers report 0 errors and 0 warnings. Two planted positives prove the check can fail (Testing below).MAINTAINERS.adoc:26says "SPDX, REUSE, OpenSSF baseline"). Known upstream test debt is disclosed in ADR 0002 and tracked upstream:main.v0.1.1,v0.2.1, andViperPkg-v0.1.1in jtv-lang) are cut at the registered commits after this merges, as ADR 0002 records.As Applicable — Documentation / Infra Changes
.machine_readable/6a2/*.a2ml: N/A. A2ML is retired estate-wide (D99/D269c), so no.a2mlfile is created or edited.TOPOLOGY.md: N/A. No architectural change.CHANGELOG.adoc.Packages.mdon the wiki still lists the old versions and theInvestigativeJournalistname. It is edited after merge, so it never advertises versions the registry does not serve.Testing
All runs used a local depot whose
registries/held General (6d0d42953fb) plus a copy of this branch. That is the same thingPkg.Registry.addproduces; in Julia:Rows marked (b156301) were re-run on this head after
mainwas merged in. The rest were measured atcf67a95. The merge brought in #109's nine Marid packages and its Layer 2b change, and touched none of the 11 entries those rows test.RegistryCI.testwithregistry_deps = [General], Julia 1.10.10 (b156301)main7c52894: 692/692). Before the merge: 553/553 (main: 550/550)ci/validate.jldefault mode, Julia 1.10.10 (b156301)expected_general_gap, and two of them are #109's MaridCodec and MaridTransport. The set is identical tomain7c52894's (670 / 22) apart from InvestigativeJournalist→InvestigativeJournalismL/LowLevel/Deps.tomldependency "HardwareResilience" has uuid d5f6a7b8… but Registry.toml binds … 01a11d29-d880…Package.tomlSiliconCore: Registry.toml uuid 01a11d29-d88c-… != Package.toml uuid c4e5f6a7-…Registry.toml+[A-Z]/*/*.toml(b156301)main7c52894: 26 hits in 14 files. This branch: 0. Only the ADR and CHANGELOG name themPkg.addof each of the 11 into a fresh env (b156301)05a805f3, its merged squash treePkg.test()against this branch, Julia 1.10.10InvestigativeJournalism.jl), 28/28 (E2E Pipeline), 870/870 (Property-Based). The tested tree is6e624d20, InvestigativeJournalism.jl#78's headea0faa0, which its Causals 0.2.1 and ZeroProb 0.1.1 dependencies resolve from this registrydiff -ris empty🤖 Generated with Claude Code
https://claude.ai/code/session_016XJoh8YA2sCWQmq93CF8RX