Stop attesting NuGet versions the writers refuse

crates/socket-patch-cli/CLI_CONTRACT.md

@@ -539,15 +539,15 @@ Discovery is read-only, never touches the network, and never fails the run: a ma
 | composer | `composer.lock` (`packages` + `packages-dev`) | `dist.url` on the patch host | `dist: {type: "path", url: ".socket/vendor/composer/<uuid>/…", reference: "<uuid>"}` | `dist.shasum`, required |
 | maven | `pom.xml` (+ `.mvn/maven.config`, `.mvn/checksums/checksums.sha256`) | a dependency version `<base>-socket.<hex8>` matching exactly ONE `socket-patch-<uuid>` repository on the patch host | pre-v5 single-POM wiring only: `socket-patch-vendor-<uuid>` repository + exactly one jar under `.socket/vendor/maven/<uuid>/` with a matching `.sha1`. A v5 suffixed pin served from `.socket/vendor/maven2` is no reference here: its ledger entry's wiring check gates it, so without a ledger it is not attested | Trusted Checksums line when enabled; not required (the suffixed version is the pin) |
 | gradle (v5.0) | `.socket/gradle/hosted-index.tsv`, the owned hosted script, and every settings script and lock file the script graph reaches | an index row whose repository URL is on the patch host and names the row's uuid, live only while the owned script is intact, every build's settings file applies it with the current index digest, every lock entry of the GA is the suffixed version, no `settings-gradle.lockfile` names the GA and no build script sets a custom `lockFile` (otherwise `patched_ref_invalid`) | none here: a vendored Gradle entry is gated by its ledger entry's wiring check | the suffixed copies installed in the Gradle cache; required (never the lock basis), because the script lets a higher upstream version resolve |
-| nuget | the first of `nuget.config` / `NuGet.config` / `NuGet.Config`, + `packages.lock.json` | source `socket-patch-<uuid>` + its exclusive exact-id `<packageSourceMapping>`; version from `packages.lock.json` | the same mapping onto `.socket/vendor/nuget/<uuid>`; version from the lock, else the feed's single nupkg | `contentHash`, required |
+| nuget | the first of `nuget.config` / `NuGet.config` / `NuGet.Config`, + `packages.lock.json` | source `socket-patch-<uuid>` + its exclusive exact-id `<packageSourceMapping>`; version from `packages.lock.json` | the same mapping onto `.socket/vendor/nuget/<uuid>`; version from the feed's single nupkg (the patched version), else the lock | `contentHash`, required. Lock versions match under `normalize_nuget_version` of the trimmed `resolved`, as the writers match them (`1.0` and `1.0.0` are one version; an empty one is `patched_ref_invalid`). When the governed locks also lock the mapped id at a version other than the patched one (another target framework or project), which both writers refuse (`redirect_nuget_lock_other_version`, `vendor_nuget_lock_other_version`), the reference is kept but not attested (`vex_nuget_lock_other_version`, naming the other versions and their locks). Vendored, the patched version is the feed's single nupkg: it is the only reference, even when no lock entry is at it, and other locked versions are never references. A hosted url names no version, so each hosted reference is checked against its own version: any two locked versions withhold them all, and a single locked version other than the patch's is gated by the patch record (`record_mismatch`). |
 | deno | `deno.lock`'s npm section, only as evidence against an npm-family pin of a `name@version` it also locks, which `vex` then omits (see **Unattested references**, #406); hosted and vendored npm runs warn about such a pin (`redirect_npm_deno_lock` / `vendor_npm_deno_lock`) | — (no hosted mode) | — (no vendored backend) | — |
 
 Recognition rules that hold for every ecosystem:
 
 * **Patch hosts.** A hosted reference counts only on `https://patch.socket.dev` or the `--patch-server-url` / `SOCKET_PATCH_SERVER_URL` origin, with no userinfo. The uuid is the URL's LAST canonical-uuid path segment, because grant tokens may themselves be uuid-shaped. The Go module prefix is fixed. `socket-patch-<uuid>` registry / repository / source names count only through a pin. For a URL on any other host, see **Patch hosts** above.
 * **Pins, not definitions.** A registry, index or source *definition* alone (cargo `[registries]`, nuget `<add>`, pom `<repository>`, uv index tables, `.npmrc`) never makes a reference, because it survives a reverted pin. Sections the package manager ignores are not read: npm's v2 `dependencies` mirror, a `.cargo/config.toml` shadowed by `.cargo/config`. A Socket pin inside a maven `<profile>` is diagnosed, never a reference.
 * **Contested locks.** When one lock wires a package to a patch and another lock resolves the same `name@version` from a non-Socket source, the build's bytes depend on which package manager runs. The reference is then dropped with a `patched_ref_unattributable` diagnostic naming both files. This applies across npm / pnpm / yarn / bun and across uv / pylock / poetry / pdm / Pipfile.lock / requirements. PEP 723 script locks neither contest nor are contested. The root `requirements.txt` and its in-root `-r` includes count as one lock here: pip reads them as one requirement set, where a direct reference wins over a compatible `==` pin of the same version in another file of the set (#1086). A PyPI version matches under PEP 440 equality, so a requirements line pinned in any spelling that can only install the wired release (`==1.2.3.0`, `==v1.2.3`, `===1.2.3`) contests it like `==1.2.3` does; a range or wildcard (`~=1.2.3`, `==1.2.*`) does not (#1602). A **bundled** npm copy (`inBundle: true`, or v1 `bundled: true`, beneath a dependency that bundles it; the root project's own bundles install from the lock, #1618) of the same `name@version` contests the reference too, in the same lock, in the other npm lock of a shrinkwrap/package-lock pair, or in any other lock. npm unpacks it from the parent package's tarball, so no rewire reaches it and it stays unpatched. Bun and vlt unpack bundled copies the same way (#469, #471). pnpm unpacks bundled copies too, but its lock cannot tie one to a reference (see **Unattested references** below). For Bun, that is a `bun.lock` entry whose meta is `{ "bundled": true }`, or a `bun.lockb` record that a dependency edge with the `bundled` behavior bit reaches, including one Bun shares with a regular install. Such an entry is never a reference, and it contests the reference the same way. For vlt, the lock records no node for a bundled copy, so the copy is found in the installed store: a real package directory inside a store package's own `node_modules`. yarn classic's `yarn.lock` records nothing about bundled copies either (#758), so for a lock with a Socket wiring they are found in the installed `node_modules`: each package a parent's `bundleDependencies` / `bundledDependencies` names inside the parent's own `node_modules`, the dependencies of those beside them, and everything nested under them. Such a copy contests the reference in `yarn.lock` (the diagnostic names the parent's lock entry and the copy's directory); a lock-only `vex` with no `node_modules` can't see it. Hosted and vendored scans skip these copies with `redirect_bun_bundled_instance_skipped` / `redirect_vlt_bundled_instance_skipped` / `vendor_bundled_instance_skipped`, and a yarn classic hosted scan names one with `redirect_copy_unpatched`. When a bundled copy is the only instance, vendoring refuses with `vendor_lock_entry_not_rewritable`. A reference withheld only because such an unreachable copy installs beside it (a bundled npm / Bun / vlt copy, an npm copy beneath a `hasShrinkwrap` package (#753) or one npm installs from a git / url / `file:` spec, or a yarn classic git or `file:` directory block of the same `name@version`) is still the rewriter's own wiring of exactly one package version: it is never attested, but `rollback`, `remove`, `list` and the hosted → vendored takeover treat it as an ordinary hosted pin and restore its upstream registry entry (#828), rather than refusing it with `hosted_wiring_contested`. Another entry of the **same** npm, Bun or yarn lock that resolves the wired `name@version` from a non-Socket source (for example a workspace member added after the rewire, then `npm install` / `bun install` / `yarn install`; for yarn berry, a registry locator such as `left-pad@npm:1.3.0` beside the hosted `…::__archiveUrl=` one) contests the reference too (#588). The package manager installs both entries, and that copy stays unpatched. Re-running `scan` / `vendor` rewires every copy.
-* **Unattested references.** Some evidence shows a build may run a copy no wiring reaches, but cannot be tied to the reference's exact `name@version` or cannot say the build runs it. The reference then stays a reference: `list`, `rollback` and `remove` find it, and the ledgers' liveness gates (`vendor --check`, `scan`) keep treating the wiring as live, because re-running `scan` / `vendor` could never clear the evidence. Only `vex` omits it, as a run warning and as the `failed[].reason`. Four cases besides Gradle's (`vex_gradle_lock_above_base`): **pnpm bundled copies** (`vex_pnpm_bundled_copy`): a `packages:` entry's `bundledDependencies:` names the copies pnpm unpacks from that package's own tarball, but pnpm never locks them, so the bundled version is not in the lock. A pnpm reference whose package name a `bundledDependencies` list in the same lock names is omitted whatever its version (a missed attestation when the bundled copy is another version, never a false one), and `bundledDependencies: true` (or a value that cannot be read) omits every reference of that lock. It reaches no other lock. **deno.lock** (`vex_deno_lock_copy`): `deno install` installs a `package.json` project's npm dependencies from `deno.lock` and never reads `package-lock.json` / `pnpm-lock.yaml` / `yarn.lock`, so an npm-family reference whose `name@version` the `deno.lock` npm section also locks is omitted (#406). Whether Deno or another package manager populates `node_modules` (`nodeModulesDir: "manual"` allows either) is not in the files, so this holds whatever `nodeModulesDir` says. **Maven reactor roots** (`vex_maven_reactor_root`): a hosted pin in a root `pom.xml` that declares `<modules>` / `<subprojects>`, or whose subprojects Maven 4 discovers (see the hosted Maven rewriter), is omitted, because a module's own literal `<version>` overrides a root pin and module poms are not read (#261); hosted scan refuses such roots (`redirect_maven_multimodule_unsupported`), so this only covers pins an older release wrote. **Maven classifier copies** (`vex_maven_classifier_unpatched`): a `<classifier>` dependency of a hosted-pinned GA is its own coordinate; the classifier-less pin does not retarget it and the grant serves only the main jar, so Maven fetches it from the next repository. When a live, non-profile classifier copy other than `sources` / `javadoc` (`tests`, a native or JDK-specific build) resolves any version other than the suffixed pin — its own literal `<version>`, a `<dependencyManagement>` entry with the same classifier, or no version in the root pom — the pin is omitted. `sources` / `javadoc` copies never run on a classpath and do not withhold the statement; hosted scan only warns about them (`redirect_maven_classifier_unsupported`) and skips the dep when an executable copy names the patched release.
+* **Unattested references.** Some evidence shows a build may run a copy no wiring reaches, but cannot be tied to the reference's exact `name@version` or cannot say the build runs it. The reference then stays a reference: `list`, `rollback` and `remove` find it, and the ledgers' liveness gates (`vendor --check`, `scan`) keep treating the wiring as live, because re-running `scan` / `vendor` could never clear the evidence. Only `vex` omits it, as a run warning and as the `failed[].reason`. Four cases besides Gradle's (`vex_gradle_lock_above_base`) and NuGet's (`vex_nuget_lock_other_version`, see the table above): **pnpm bundled copies** (`vex_pnpm_bundled_copy`): a `packages:` entry's `bundledDependencies:` names the copies pnpm unpacks from that package's own tarball, but pnpm never locks them, so the bundled version is not in the lock. A pnpm reference whose package name a `bundledDependencies` list in the same lock names is omitted whatever its version (a missed attestation when the bundled copy is another version, never a false one), and `bundledDependencies: true` (or a value that cannot be read) omits every reference of that lock. It reaches no other lock. **deno.lock** (`vex_deno_lock_copy`): `deno install` installs a `package.json` project's npm dependencies from `deno.lock` and never reads `package-lock.json` / `pnpm-lock.yaml` / `yarn.lock`, so an npm-family reference whose `name@version` the `deno.lock` npm section also locks is omitted (#406). Whether Deno or another package manager populates `node_modules` (`nodeModulesDir: "manual"` allows either) is not in the files, so this holds whatever `nodeModulesDir` says. **Maven reactor roots** (`vex_maven_reactor_root`): a hosted pin in a root `pom.xml` that declares `<modules>` / `<subprojects>`, or whose subprojects Maven 4 discovers (see the hosted Maven rewriter), is omitted, because a module's own literal `<version>` overrides a root pin and module poms are not read (#261); hosted scan refuses such roots (`redirect_maven_multimodule_unsupported`), so this only covers pins an older release wrote. **Maven classifier copies** (`vex_maven_classifier_unpatched`): a `<classifier>` dependency of a hosted-pinned GA is its own coordinate; the classifier-less pin does not retarget it and the grant serves only the main jar, so Maven fetches it from the next repository. When a live, non-profile classifier copy other than `sources` / `javadoc` (`tests`, a native or JDK-specific build) resolves any version other than the suffixed pin — its own literal `<version>`, a `<dependencyManagement>` entry with the same classifier, or no version in the root pom — the pin is omitted. `sources` / `javadoc` copies never run on a classpath and do not withhold the statement; hosted scan only warns about them (`redirect_maven_classifier_unsupported`) and skips the dep when an executable copy names the patched release.
 * **Shrinkwrap without a twin (#899).** npm >= 12 never reads `npm-shrinkwrap.json`: on a project whose only npm lock is the shrinkwrap it resolves the tree from the registry and writes a fresh `package-lock.json`, so a Socket reference there reaches npm <= 11 only. The reference is still discovered, so `list`, `rollback` and `remove` manage it, but `vex` omits the patch (`vex_npm_shrinkwrap_only`, as a run warning and as the `failed[].reason`). A `package-lock.json` beside it (wired the same way) makes it attestable again. Hosted and vendored runs still rewire the shrinkwrap and warn `redirect_npm_shrinkwrap_only` / `vendor_npm_shrinkwrap_only`.
 * **Lockless pins.** With no lock to name a version, a `Cargo.toml` pin (every declaration on `socket-patch-<uuid>`, that registry defined on the patch host for the same uuid) or an exclusive nuget exact-id mapping is never a reference on its own, so v5.0 does not attest it (nor does `list` show it, or `rollback` / `remove` restore it — restore those files from version control). Only a pre-v5 redirect-ledger record naming a version the pin admits keeps it live. The same holds for a gem wired only in the `Gemfile` (the pre-bundler-2.6 mixed state, lock not converged).
 
@@ -2678,6 +2678,7 @@ match it, withholds the statement:
 | `vex_deno_lock_copy` | run warning and `failed[].reason` | An npm pin is wired, but `deno.lock` locks the same `name@version`, which `deno install` installs without the wiring; no statement while it does. `vendor --check` and `scan` still treat the wiring as live. |
 | `vex_maven_reactor_root` | run warning and `failed[].reason` | A hosted Maven pin sits in a root `pom.xml` that declares `<modules>` / `<subprojects>` (a profile's included), where a module's own literal `<version>` overrides it; only the root pom is read, so no statement (#261). `list`, `rollback` and `remove` still find the pin. |
 | `vex_maven_classifier_unpatched` | run warning and `failed[].reason` | A hosted Maven pin's GA also has a live executable `<classifier>` copy (not `sources` / `javadoc`) at a version other than the pin, which the classifier-less pin does not retarget and the grant does not serve; no statement while it does. `list`, `rollback` and `remove` still find the pin. |
+| `vex_nuget_lock_other_version` | run warning and `failed[].reason` | A NuGet Socket source maps an id that a governed `packages.lock.json` also locks at a version other than the patched one; the mapping routes every version to a source that serves only the patched one, so that target framework cannot restore, and both writers refuse such a lock. No statement until every target framework locks only the patched version. `list`, `rollback` and `remove` still find the wiring. |
 | `vex_gradle_derived_cache_unchecked` | run warning | The derived-cache walk was cut short (a very large transforms cache); the statement is still emitted, the copies the walk did reach were checked. |
 
 A vendored Gradle entry attests only while its wiring is live (apply line, index rows,

crates/socket-patch-cli/src/commands/vex_sources.rs

@@ -232,6 +232,10 @@ const NOTE_MAVEN_REACTOR_ROOT: &str = "vex_maven_reactor_root";
 /// classifier copy at another version, which the pin does not reach
 /// (`vex::Unattested`).
 const NOTE_MAVEN_CLASSIFIER_UNPATCHED: &str = "vex_maven_classifier_unpatched";
+/// Omission tag and note: a NuGet Socket source's id is also locked at a
+/// version other than the patched one, which that source cannot serve
+/// (`vex::Unattested`).
+const NOTE_NUGET_LOCK_OTHER_VERSION: &str = "vex_nuget_lock_other_version";
 
 /// The omission tag of an [`Unattested`](socket_patch_core::vex::Unattested)
 /// ref, and the remedy its note ends with.
@@ -268,6 +272,11 @@ fn unattested_note(kind: UnattestedKind) -> (&'static str, &'static str) {
             "not attested while that classifier copy resolves an unpatched version; the \
              wiring itself is intact, so re-running `scan` does not change this",
         ),
+        UnattestedKind::NugetLockOtherVersion => (
+            NOTE_NUGET_LOCK_OTHER_VERSION,
+            "not attested until every target framework locks only the patched version (align \
+             them on it, restore, and re-run `scan` / `vendor`)",
+        ),
     }
 }
 

crates/socket-patch-core/src/vex/discover/mod.rs

@@ -606,6 +606,13 @@ pub enum UnattestedKind {
     /// version: the classifier-less pin does not retarget it and the grant
     /// serves only the main jar, so the build runs the public classifier jar.
     MavenClassifierUnpatched,
+    /// A NuGet Socket source whose mapped id a governed `packages.lock.json`
+    /// also locks at a version other than the patched one: the mapping
+    /// routes every version of the id to a source that serves only the
+    /// patched one, so that target framework cannot restore — the case both
+    /// writers refuse (`redirect_nuget_lock_other_version`,
+    /// `vendor_nuget_lock_other_version`).
+    NugetLockOtherVersion,
 }
 
 /// A ref discovery emits (so rollback, remove and list find the wiring,