Compel agents to keep acting until they explicitly finish
Forcing only the opening turn let a weak model take one token action and stop, so a kickoff hired one agent (or none) and went quiet. Now every turn is forced (tool_choice required) and the agent ends only by calling the new `finish` tool — so it must keep hiring, delegating, and messaging until the work is actually done. Finishing with zero actions, or returning no tool call at all, fails the run loudly instead of faking success. Prompts tell each agent it must call a tool every turn and finish only when done. Tests drive weak-model stubs that act then finish, proving a full team forms rather than a single gesture. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
1 parent
ef9ee463c9
commit
f7543e506d
7 files changed
+72
-50
No files matched your search
@@ -37,10 +37,11 @@ describe("task execution service", () => {
|
||||
const fakeToolTurn = async (input: ProviderToolTurnInput): Promise<AgentTurnResult> => {
|
||||
const first = input.messages[0];
|
||||
if (first?.role === "user") capturedPrompt = first.text;
|
||||
if (input.toolChoice === "required") {
|
||||
return { text: "executed:gpt-5.4", toolCalls: [{ id: `call_${randomUUID()}`, name: "create_document", arguments: { title: "Run note", doc_type: "memo", content: "Executed." } }], rawResponseJson: JSON.stringify({ id: "resp-task-1a" }) };
|
||||
}
|
||||
return { text: "executed:gpt-5.4", toolCalls: [], rawResponseJson: JSON.stringify({ id: "resp-task-1b" }) };
|
||||
const acted = input.messages.some((message) => message.role === "tool");
|
||||
const toolCall = acted
|
||||
? { id: `call_${randomUUID()}`, name: "finish", arguments: {} }
|
||||
: { id: `call_${randomUUID()}`, name: "create_document", arguments: { title: "Run note", doc_type: "memo", content: "Executed." } };
|
||||
return { text: "executed:gpt-5.4", toolCalls: [toolCall], rawResponseJson: JSON.stringify({ id: "resp-task-1" }) };
|
||||
};
|
||||
const providers = new ProviderService(connection.db, providerAdapters, testVault, undefined, fakeToolTurn);
|
||||
const skills = new SkillService(connection.db, join(tmpdir(), `agentcompany-task-skills-${randomUUID()}`));
|
||||
@@ -145,10 +146,11 @@ describe("task execution service", () => {
|
||||
const approvals = new ApprovalService(connection.db);
|
||||
const snapshot = new SnapshotService(connection.db);
|
||||
const fakeToolTurn = async (input: ProviderToolTurnInput): Promise<AgentTurnResult> => {
|
||||
if (input.toolChoice === "required") {
|
||||
return { text: "executed:gpt-5.4", toolCalls: [{ id: `call_${randomUUID()}`, name: "create_document", arguments: { title: "Run note", doc_type: "memo", content: "Executed." } }], rawResponseJson: JSON.stringify({ id: "resp-task-approved-a" }) };
|
||||
}
|
||||
return { text: "executed:gpt-5.4", toolCalls: [], rawResponseJson: JSON.stringify({ id: "resp-task-approved-b" }) };
|
||||
const acted = input.messages.some((message) => message.role === "tool");
|
||||
const toolCall = acted
|
||||
? { id: `call_${randomUUID()}`, name: "finish", arguments: {} }
|
||||
: { id: `call_${randomUUID()}`, name: "create_document", arguments: { title: "Run note", doc_type: "memo", content: "Executed." } };
|
||||
return { text: "executed:gpt-5.4", toolCalls: [toolCall], rawResponseJson: JSON.stringify({ id: "resp-task-approved" }) };
|
||||
};
|
||||
const providers = new ProviderService(connection.db, providerAdapters, testVault, undefined, fakeToolTurn);
|
||||
const execution = new TaskExecutionService(connection.db, providers);
|
||||
|
||||
Reference in new issue
Block a user