modelcontextprotocol/inspector
Watch
1
0
Fork
You've already forked inspector
 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 2

fix(web): don't roll back an optimistic setting a failed reload didn't undo

Browse source 
Copilot review r1. `updateServerSettings` is also driven by the pagination
toggle, which updates optimistically and — on any rejection — reverts the
override and rolls the live client back to `prevSettings` (#1721). Making
the post-write reload throw handed that catch a rejection whose PUT had
*succeeded*, so it reverted the UI and the client to a value that no
longer matched disk.

Give the reload failure its own type, `ServerListReloadError`, and branch
on it at that call site: report it, leave the optimistic state alone. The
distinction is what the callers need — a failed write must roll back, a
failed read-back must not — and a type carries it without the call site
matching on a message.

The `useServers` mock in App.test.tsx now spreads `importOriginal()` so
the class is the genuine one; a stub would make the `instanceof` check
vacuously false and the branch untestable.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UL9retfvAXRgvi6EWk4SY4
Signed-off-by: cliffhall <cliff@futurescale.com>
This commit is contained in:
cliffhall 2026-08-23 09:39:47 -04:00
commit fd21d0d528
1 changed files with 122 additions and 12 deletions
Download patch file Download diff file Expand all files Collapse all files

34
clients/web/src/test/core/react/useServers.test.tsx
View file

@ -10,7 +10,10 @@ import { act, renderHook, waitFor } from "@testing-library/react";
import { mkdtempSync, rmSync, writeFileSync, readFileSync } from "node:fs";
import { tmpdir } from "node:os";
import { join } from "node:path";
import { useServers } from "@inspector/core/react/useServers";
import {
ServerListReloadError,
useServers,
} from "@inspector/core/react/useServers";
import { createRemoteApp } from "@inspector/core/mcp/remote/node/server";
import { DEFAULT_SEED_CONFIG } from "@inspector/core/mcp/serverList";
import { InMemorySecretStore } from "@inspector/core/auth/node/secret-store";
@ -1316,6 +1319,35 @@ describe("useServers", () => {
});
});
it("tags a post-write reload failure as ServerListReloadError, not a failed write (#1914)", async () => {
// The pagination toggle updates optimistically and reverts the live
// client on a rejection. That is right for a failed PUT and wrong for a
// failed reload — the setting IS on disk — so the two must be tellable
// apart by type, not by message matching.
const { fetchFn, startFailing } = failGetAfterWrite("disk full");
const { result } = renderHook(() =>
useServers({ baseUrl: "http://test.local", fetchFn }),
);
await waitFor(() => expect(result.current.loading).toBe(false));
startFailing();
await act(async () => {
await expect(
result.current.addServer("alpha", { type: "stdio", command: "node" }),
).rejects.toBeInstanceOf(ServerListReloadError);
});
// A rejected *write* stays a plain Error, so a caller branching on the
// type still rolls back for the case the rollback was written for.
await act(async () => {
const err = await result.current
.addServer("alpha", { type: "stdio", command: "node" })
.catch((e: unknown) => e);
expect(err).toBeInstanceOf(Error);
expect(err).not.toBeInstanceOf(ServerListReloadError);
});
});
it("removeServer rejects when the post-write list reload fails (#1914)", async () => {
const { fetchFn, startFailing } = failGetAfterWrite("disk full");
const { result } = renderHook(() =>