diff options
| author | Adam Malczewski <[email protected]> | 2026-06-02 15:53:15 +0900 |
|---|---|---|
| committer | Adam Malczewski <[email protected]> | 2026-06-02 15:53:15 +0900 |
| commit | aa295e82197ebc77d9466eee28380bc5bcc0863d (patch) | |
| tree | 2a900e1f10564f4f4b3f698d1058ebc37d5c8d45 /packages/core/tests | |
| parent | e475e527cd768dc05368a0881a07a84ea140e13e (diff) | |
| download | dispatch-aa295e82197ebc77d9466eee28380bc5bcc0863d.tar.gz dispatch-aa295e82197ebc77d9466eee28380bc5bcc0863d.zip | |
fix(tabs): only mention read_tab when the sender actually has it; CAPS on ONLY
The send_to_tab guidance previously told the agent it could call read_tab to
check for a reply, but the tab-messaging permissions are split — a tab can
hold send_to_tab WITHOUT read_tab (the exact case in testing). Advertising a
tool the agent wasn't granted is wrong.
Thread a canReadTab flag from AgentManager.buildTabCommToolEntries into
createSendToTabTool (true iff this tab is also granted read_tab). The tool
description and the delivery-result text now only reference read_tab when
canReadTab is true; otherwise they say a reply arrives on its own and to end
the turn. Drop the read_tab phrasing from the static TOOL_DESCRIPTIONS
one-liner (can't be conditional per-tab there).
Also uppercase ONLY in the recipient reply-contract footer for emphasis.
Tests: cover both canReadTab branches for description + result text; assert
ONLY is uppercased.
Diffstat (limited to 'packages/core/tests')
| -rw-r--r-- | packages/core/tests/tools/send-to-tab.test.ts | 33 |
1 files changed, 32 insertions, 1 deletions
diff --git a/packages/core/tests/tools/send-to-tab.test.ts b/packages/core/tests/tools/send-to-tab.test.ts index 68f8fa0..48ff460 100644 --- a/packages/core/tests/tools/send-to-tab.test.ts +++ b/packages/core/tests/tools/send-to-tab.test.ts @@ -14,6 +14,7 @@ function makeCallbacks(overrides: Partial<SendToTabCallbacks> = {}): SendToTabCa deliver: () => ({ status: "started" }), listOpenHandles: () => [{ handle: "targ", title: "Target" }], self: { id: "self-id", handle: "self" }, + canReadTab: true, ...overrides, }; } @@ -28,6 +29,19 @@ describe("createSendToTabTool — schema & description", () => { expect(tool.description.toLowerCase()).toContain("do not sleep"); expect(tool.description.toLowerCase()).toContain("end your turn"); }); + + it("mentions read_tab in the description only when canReadTab is true", () => { + const tool = createSendToTabTool(makeCallbacks({ canReadTab: true })); + expect(tool.description).toContain("read_tab"); + }); + + it("never mentions read_tab in the description when canReadTab is false", () => { + const tool = createSendToTabTool(makeCallbacks({ canReadTab: false })); + expect(tool.description).not.toContain("read_tab"); + // Still tells the agent a reply arrives on its own + to end its turn. + expect(tool.description.toLowerCase()).toContain("arrives on its own"); + expect(tool.description.toLowerCase()).toContain("end your turn"); + }); }); describe("createSendToTabTool — execute()", () => { @@ -46,7 +60,7 @@ describe("createSendToTabTool — execute()", () => { // Reply contract: the recipient must answer via send_to_tab back to the // sender's handle, not as a plain text reply to its own user. expect(delivered).toContain('send_to_tab tool with tab_id "self"'); - expect(delivered.toLowerCase()).toContain("only reply if"); + expect(delivered).toContain("ONLY reply if"); expect(out).toContain("idle"); expect(out).toContain("targ"); // Sender is steered away from busy-waiting and told to end its turn. @@ -54,6 +68,23 @@ describe("createSendToTabTool — execute()", () => { expect(out.toLowerCase()).toContain("end your turn"); }); + it("points the sender at read_tab in the result only when canReadTab is true", async () => { + const deliver = vi.fn(() => ({ status: "started" as const })); + const tool = createSendToTabTool(makeCallbacks({ deliver, canReadTab: true })); + const out = await tool.execute({ tab_id: "targ", message: "hi" }); + expect(out).toContain("read_tab"); + }); + + it("omits read_tab from the result when canReadTab is false", async () => { + const deliver = vi.fn(() => ({ status: "started" as const })); + const tool = createSendToTabTool(makeCallbacks({ deliver, canReadTab: false })); + const out = await tool.execute({ tab_id: "targ", message: "hi" }); + expect(out).not.toContain("read_tab"); + // Still steers away from busy-waiting and toward ending the turn. + expect(out.toLowerCase()).toContain("do not sleep"); + expect(out.toLowerCase()).toContain("end your turn"); + }); + it("reports the queued status when the target is busy", async () => { const deliver = vi.fn(() => ({ status: "queued" as const })); const tool = createSendToTabTool(makeCallbacks({ deliver })); |
