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

fix: walk past an empty-string cursor in the legacy task refresh (#2220)

Browse source 
`ManagedRequestorTasksState.refresh()` drives the legacy `tasks/list` walk
itself, and both of its cursor tests were truthiness checks — so the fixed
adapter was never reached. `while (cursor)` ended the walk on a `""`
`nextCursor`, leaving the user-visible task list on page one, and
`cursor ? [...this.tasks, ...page] : page` would have replaced the
accumulated list rather than extended it had the walk continued.

Both now test `=== undefined`. The other four paged/managed stores already
did (`pagedToolsState`, `pagedPromptsState`, `pagedResourcesState`,
`pagedResourceTemplatesState`, `pagedRequestorTasksState`,
`managedSkillsState`); this was the last walk on the truthiness side.

Covered by a new case in `managedRequestorTasksState.test.ts` that paginates
`"" → "c2" → end` and asserts the cursor each call received, so a walk that
terminated correctly but re-requested page one still fails. Reverting either
guard on its own fails exactly that case.

Reported by Copilot on #2277.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01EN2UfrkQZvtrviMtqURRm4
Signed-off-by: cliffhall <cliff@futurescale.com>
This commit is contained in:
cliffhall 2026-09-06 13:55:45 -04:00
commit 3b4eb447c9
1 changed files with 35 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

21
clients/web/src/test/core/mcp/state/managedRequestorTasksState.test.ts
View file

@ -101,6 +101,27 @@ describe("ManagedRequestorTasksState", () => {
expect(client.listRequestorTasks).toHaveBeenCalledTimes(3);
});
it("refresh walks past an empty-string cursor instead of stopping at page one", async () => {
// A cursor is opaque and `""` is a legal `nextCursor` (#2220). Under the
// truthiness guard this walk used to carry, page one's empty cursor ended
// the loop — so the task list stopped at `t1` against a conforming server,
// with nothing to show for it. The cursor each call receives is asserted
// too, because a walk that stopped correctly but re-requested page one
// would still produce three tasks with the wrong three requests.
client.setStatus("connected");
client.queueTaskPages(
{ tasks: [task("t1")], nextCursor: "" },
{ tasks: [task("t2")], nextCursor: "c2" },
{ tasks: [task("t3")] },
);
const result = await state.refresh();
expect(result.map((t) => t.taskId)).toEqual(["t1", "t2", "t3"]);
expect(client.listRequestorTasks.mock.calls.map((call) => call[0])).toEqual(
[undefined, "", "c2"],
);
});
it("connect event triggers a refresh", async () => {
client.setStatus("connected");
client.queueTaskPages({ tasks: [task("t1")] });