fix: run the provider sandbox and admit a release to a named environment

The provider sandbox never ran. bubblewrap 0.9.0 stops parsing an `--args`
file at the first non-option and never hands the remainder back, so the
command written into that file was silently dropped: bwrap printed its usage
text, exited 1, and the provider produced no evidence at all. The options
still travel in the args file — that is what keeps host paths and credentials
out of `/proc/<pid>/cmdline` — but the command now rides on real argv, and
`encodeProviderBwrapInput` refuses a `--` so the drop cannot come back.

The scope wrapper then could not exit. It read the supervisor's liveness pipe
through `fs`, which runs a blocking `read(2)` on a threadpool thread; the
supervisor holds that pipe open for the scope's whole life, so the read never
returned and closing the descriptor did not interrupt it. Once bubblewrap
finished the wrapper deadlocked in `process.exit`, the scope outlived the
provider, and a completed run was reported as a timeout kill. The channel is
now read through the event loop, so teardown is observable and terminal.

Creation modes were left to the ambient umask. `mkdir(mode)` and `open(mode)`
are requests the kernel subtracts the umask from, so a runner exporting a
restrictive umask produced directories it could not enter and handed `tar` a
file it could not re-open. Private modes are pinned instead of inherited.

Promotion cleanup deleted before it checked. Removals run through a pinned
descriptor, so a leaf substituted after validation had this promotion's exact
five destroyed first and the substitution reported afterwards, leaving a
half-emptied directory a retry could not tell from a completed one. The name
is re-bound to the inode before anything is removed, so the failure is total.

Separately, release coherence proved the artifacts agreed with each other but
never that they belonged where they were going: a build whose runtime document
said `APP_ENV: local`, `AUTH_MODE: demo` and a loopback API is coherent with
itself and passed every gate. `public/` is copied verbatim into `dist/`, so
that local document shipped with every build regardless of what the build was
for. Runtime configuration now comes from a declared profile, and FE-GATE-027
refuses to admit an artifact to an environment it does not match — including
refusing an undeclared destination, so nothing is admitted by omission.

`REQUEST_TIMEOUT_MS` and `VITE_ROUTER_BASE_PATH` were validated and then
dropped: the V3 executor ran every operation on its contract's own deadline,
and Vite emitted root-absolute assets for a sub-path deployment. The timeout is
now a ceiling that may tighten a contract but never loosen one, and one base
path feeds the router, the Service Worker scope and the asset base together.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This commit is contained in:
DongHyeonka
2026-08-15 16:38:19 +09:00
co-authored by Claude Opus 5
parent a0fbafb77b
commit dfb7734674
28 changed files with 1082 additions and 48 deletions
+30 -5
View File
@@ -842,12 +842,15 @@ describe("candidate archive and provider upload boundaries", () => {
expect((await readdir(path.dirname(sealedPath))).filter((leaf) => leaf.includes(".guardian-")))
.toEqual([]);
// `providerWriter` is only a path; the retry has to materialise the script
// it names or the clean-retry claim is proven by a module-not-found error.
await writeFile(fixture.providerWriter, providerV2WriterSource());
const retried = runProviderSupervisor(fixture, {
command: `node ${JSON.stringify(fixture.providerWriter)}`,
sealedPath,
});
expect(retried.status, retried.stderr).toBe(0);
}, 15_000);
}, 20_000);
it("collects the whole provider scope when its supervisor dies", async () => {
const fixture = await createProviderFixture();
@@ -1272,9 +1275,17 @@ describe("verified promotion finalizer", () => {
"provider-verification.json": valid["promotion-verification.json"],
"promotion-verification.json": valid["provider-verification.json"],
};
// `{}` never reaches the digest comparison: it fails the report schema
// first, so this case asserted a decode error while claiming to cover the
// digest branch. The substitution has to be a structurally valid report
// that simply is not the one the verification records committed to.
const divergentReport = JSON.parse(
valid["vulnerability-report.json"].toString("utf8"),
) as Record<string, any>;
divergentReport.provider = "divergent-provider";
const reportMismatch = {
...valid,
"vulnerability-report.json": Buffer.from("{}\n"),
"vulnerability-report.json": jsonBytes(divergentReport),
};
const absent = { ...valid } as Partial<typeof valid>;
delete absent["provider-verification.json"];
@@ -1513,7 +1524,10 @@ describe("verified promotion finalizer", () => {
}),
).rejects.toThrow(/leaf.*identity|staging leaf/u);
await expect(readFile(canary, "utf8")).resolves.toBe("unchanged\n");
await expect(readdir(saved)).resolves.toEqual([]);
// A substituted leaf aborts the cleanup before anything is removed, so the
// promotion this call owned is still intact and a retry sees a coherent
// directory rather than a half-emptied one.
expect((await readdir(saved)).sort()).toEqual([...PROMOTED_FILE_NAMES].sort());
}, 30_000);
it("never deletes an unrelated leaf substituted after cleanup validation", async () => {
@@ -1681,8 +1695,19 @@ async function readCgroupPids(cgroupRoot: string): Promise<number[]> {
async function waitForDirectProviderChildren(supervisorPid: number): Promise<number[]> {
for (let attempt = 0; attempt < 120; attempt += 1) {
const childrenPath = `/proc/${supervisorPid}/task/${supervisorPid}/children`;
const children = (await readFile(childrenPath, "utf8"))
.trim().split(/\s+/u).filter(Boolean).map(Number);
let listing: string;
try {
listing = await readFile(childrenPath, "utf8");
} catch (error) {
if (hasErrorCode(error, "ENOENT")) {
throw new Error(
"provider supervisor exited before its children could be observed",
{ cause: error },
);
}
throw error;
}
const children = listing.trim().split(/\s+/u).filter(Boolean).map(Number);
const arguments_ = children.flatMap((pid) => {
try {
return [showProcessArguments(pid)];