Jesse_ChenandClaude Opus 5.5
84eb23152a
T2 of TASK-rectification-grounding-20260927 (product decision P1). The rectification 「重试（重新生成）」 was a blind agent.generate rewrite that persisted unchecked text (invented range / fit rate reproduced). Removed: the regenerate route, regenerate-turn.ts, the regeneration agent and its read-only tool set, the client regenerate action/state and the regenerating/canRegenerate props. ChatMessageActions renders the regenerate button only when onRegenerate is passed; ordinary consultation is unchanged. The DB function regenerate_agentic_rectification_turn is kept (AGENTS §7.6, retire in a later round). DESIGN.md / VOICE.md updated; source-contract tests follow with 原值/新值/原因 notes. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_017eEAG8HD3mm8gsKXgk8uU8
1002 lines
45 KiB
TypeScript
1002 lines
45 KiB
TypeScript
import assert from "node:assert/strict";
|
||
import { readFileSync, realpathSync } from "node:fs";
|
||
import { basename } from "node:path";
|
||
import { fileURLToPath } from "node:url";
|
||
import test from "node:test";
|
||
|
||
import {
|
||
RECTIFICATION_AGENT_HARD_STEP_LIMIT,
|
||
RECTIFICATION_AGENT_MAX_STEPS,
|
||
RECTIFICATION_AGENT_STEP_BUDGETS,
|
||
RECTIFICATION_V9_PACKAGE_PATH,
|
||
RECTIFICATION_V9_SKILL_PATH,
|
||
RECTIFICATION_V9_SKILL_NAME,
|
||
getRectificationV9Agent,
|
||
resolveRectificationStepBudget,
|
||
} from "../src/mastra/agentic-rectification.ts";
|
||
import * as agenticRectification from "../src/mastra/agentic-rectification.ts";
|
||
import * as rectificationV9Tools from "../src/mastra/rectification-v9-tools.ts";
|
||
import {
|
||
createRectificationV9Tools,
|
||
} from "../src/mastra/rectification-v9-tools.ts";
|
||
import { buildOpeningBrief, runV9AgentTurn, type V9AgentRunOptions } from "../src/lib/rectification-agentic/v9/agent-run.ts";
|
||
import { parseV9CaseDossier, persistV9Candidate } from "../src/lib/rectification-agentic/v9/tool-service.ts";
|
||
import { RECTIFICATION_SKILL_NAME, RECTIFICATION_SKILL_VERSION } from "../src/lib/rectification-agentic/v9/case-status.ts";
|
||
import { composeCollectSpokenAssistantText } from "../src/lib/rectification-agentic/v9/collect-prompt.ts";
|
||
import {
|
||
GENERIC_COLLECT_QUESTION,
|
||
OPENING_COLLECT_DOMAINS,
|
||
RECTIFICATION_USER_COPY,
|
||
isAcceptableOpeningBody,
|
||
openingSpokenBody,
|
||
} from "../src/lib/rectification-agentic/user-copy.ts";
|
||
import {
|
||
CASE_ID,
|
||
CANDIDATE_RANGE,
|
||
CANDIDATE_ID,
|
||
EVIDENCE_ID,
|
||
SESSION_ID,
|
||
TURN_ID,
|
||
USER_ID,
|
||
activeFocusFixture,
|
||
candidateSnapshotFixture,
|
||
conversationSummaryFixture,
|
||
dossierFixture,
|
||
fakeAccounting,
|
||
receiptHandlers,
|
||
RESULT_ID,
|
||
computeFixture,
|
||
} from "./rectification-v9-test-support.ts";
|
||
|
||
const agentSource = readFileSync(
|
||
fileURLToPath(new URL("../src/mastra/agentic-rectification.ts", import.meta.url)),
|
||
"utf8",
|
||
);
|
||
|
||
test("system prompt carries only high-priority boundaries, never the method copy", () => {
|
||
const promptStart = agentSource.indexOf("const agenticRectificationInstructions");
|
||
const promptEnd = agentSource.indexOf("export function getRectificationV9Agent");
|
||
const prompt = agentSource.slice(promptStart, promptEnd);
|
||
assert.doesNotMatch(prompt, /rectification-gate[\s\S]*rectification-scan/);
|
||
assert.doesNotMatch(prompt, /rectification-score[\s\S]*rectification-diagnostics/);
|
||
assert.doesNotMatch(prompt, /rectification-confirm[\s\S]*rectification-save-birth-time/);
|
||
assert.doesNotMatch(prompt, /10[–-]15 个事件/);
|
||
assert.doesNotMatch(prompt, /run the required gate/);
|
||
assert.doesNotMatch(prompt, /candidate_range/);
|
||
assert.match(prompt, /jyotish-birth-time-rectification/);
|
||
assert.match(prompt, /display_date_label/);
|
||
assert.match(prompt, /rectification-record-evidence-batch/);
|
||
assert.match(prompt, /不可分区间/);
|
||
assert.match(prompt, /skill_verification_report/);
|
||
assert.match(prompt, /80%\/60%/);
|
||
assert.match(prompt, /confirmation_allowed/);
|
||
assert.doesNotMatch(prompt, /本会话以代表性时间收口|本轮校正已收口/);
|
||
assert.match(prompt, /工具执行保持静默/);
|
||
assert.match(prompt, /思考用简体中文写在思维链/);
|
||
assert.match(prompt, /对用户说的话必须自己写在正文里/);
|
||
assert.match(prompt, /用 rectification-set-focus 的 spokenPrompt 写出服务端给你的下一问/);
|
||
// 旧：每轮正文 2-4 句 → 新：证据轮正文只写一句复述 → BUG-606 决策 3
|
||
assert.match(prompt, /证据轮正文只写一句复述/);
|
||
assert.match(prompt, /「先这样」由服务器/);
|
||
assert.match(prompt, /职业题只问平时做什么/);
|
||
assert.match(prompt, /焦点 domain 是 career 不是 occupation/);
|
||
assert.match(prompt, /collection_progress/);
|
||
assert.match(prompt, /不得写「范围在收窄」/);
|
||
assert.doesNotMatch(prompt, /不得询问外貌、体质、胎记或疤痕/);
|
||
assert.doesNotMatch(prompt, /财务与健康只有用户主动说才问/);
|
||
assert.doesNotMatch(prompt, /id 不是 adopt_representative/);
|
||
assert.doesNotMatch(prompt, /不要调用 rectification-set-focus/);
|
||
assert.doesNotMatch(prompt, /自己写一句自然语言追问/);
|
||
assert.doesNotMatch(prompt, /运行器会把口语接到这句题干/);
|
||
assert.doesNotMatch(prompt, /外貌、体质、胎记或疤痕可以问/);
|
||
assert.doesNotMatch(prompt, /分盘句和宫位表由界面展示/);
|
||
assert.doesNotMatch(prompt, /不是整张宫位表/);
|
||
assert.doesNotMatch(prompt, /分别 propose\+confirm/);
|
||
assert.ok(prompt.split("\n").length <= 30, "instructions must stay bounded");
|
||
});
|
||
|
||
test("agent pins the dedicated rectification skill and its fixed version", () => {
|
||
assert.equal(RECTIFICATION_V9_SKILL_NAME, "jyotish-birth-time-rectification");
|
||
assert.equal(basename(RECTIFICATION_V9_SKILL_PATH), RECTIFICATION_V9_SKILL_NAME);
|
||
// 原值: "10.0.25"
|
||
// 原值: "10.0.26"
|
||
// 新值: "10.0.27"
|
||
// 原因: 出卡精度门槛 + 引导式补经历
|
||
// 原因: BUG-668 定向补事第四选项写进 Skill
|
||
// 原值: "10.0.28" / 新值: "10.0.29" / 原因: 年月阶段改口述并禁止追问本人
|
||
// 原值: "10.0.29" / 新值: "10.0.30" / 原因: D2 出卡句与 D4 卡上百分比规则写进 Skill（2026-09-26）
|
||
// 原值: "10.0.30" / 新值: "10.0.31" / 原因: 开场改大白话、题干带例子和示例回答写进 Skill OpeningPolicy（BUG-1049，2026-09-26）
|
||
assert.ok(RECTIFICATION_V9_PACKAGE_PATH.endsWith("skills/jyotish-birth-time-rectification/versions/10.0.31"));
|
||
assert.notEqual(RECTIFICATION_V9_SKILL_PATH, RECTIFICATION_V9_PACKAGE_PATH);
|
||
assert.equal(realpathSync(RECTIFICATION_V9_SKILL_PATH), RECTIFICATION_V9_PACKAGE_PATH);
|
||
assert.equal(RECTIFICATION_SKILL_NAME, "jyotish-birth-time-rectification");
|
||
// 原值: "10.0.25"
|
||
// 原值: "10.0.26"
|
||
// 新值: "10.0.27"
|
||
// 原因: 出卡精度门槛 + 引导式补经历
|
||
// 原因: BUG-668 定向补事第四选项写进 Skill
|
||
// 原值: "10.0.28" / 新值: "10.0.29" / 原因: 年月阶段改口述并禁止追问本人
|
||
// 原值: "10.0.29" / 新值: "10.0.30" / 原因: D2 出卡句与 D4 卡上百分比规则写进 Skill（2026-09-26）
|
||
// 原值: "10.0.30" / 新值: "10.0.31" / 原因: 开场改大白话、题干带例子和示例回答写进 Skill OpeningPolicy（BUG-1049，2026-09-26）
|
||
assert.equal(RECTIFICATION_SKILL_VERSION, "10.0.31");
|
||
});
|
||
|
||
test("step budgets are bounded per action with a hard ceiling", () => {
|
||
assert.equal(RECTIFICATION_AGENT_STEP_BUDGETS.opening, 6);
|
||
assert.equal(RECTIFICATION_AGENT_STEP_BUDGETS.evidence, 8);
|
||
assert.equal(RECTIFICATION_AGENT_STEP_BUDGETS.rescore, 12);
|
||
assert.equal(RECTIFICATION_AGENT_STEP_BUDGETS.accept, 6);
|
||
assert.equal(RECTIFICATION_AGENT_MAX_STEPS, 12);
|
||
for (const action of ["opening", "read_only", "evidence", "rescore", "accept", "confirm"]) {
|
||
const budget = resolveRectificationStepBudget(action as keyof typeof RECTIFICATION_AGENT_STEP_BUDGETS);
|
||
assert.ok(budget <= RECTIFICATION_AGENT_HARD_STEP_LIMIT, `${action} must respect the hard ceiling`);
|
||
}
|
||
assert.equal(resolveRectificationStepBudget("rescore"), 12);
|
||
});
|
||
|
||
test("no tool accepts event arrays, birth data or candidate ranges as input", () => {
|
||
const tools = createRectificationV9Tools({
|
||
userId: USER_ID,
|
||
caseId: CASE_ID,
|
||
turnId: TURN_ID,
|
||
accounting: fakeAccounting({}).client as never,
|
||
});
|
||
const toolNames = [
|
||
"rectification-read-case",
|
||
"rectification-propose-evidence",
|
||
"rectification-confirm-evidence",
|
||
"rectification-revise-evidence",
|
||
"rectification-compare-candidates",
|
||
"rectification-read-diagnostics",
|
||
"rectification-offer-candidates",
|
||
"rectification-accept-candidate",
|
||
"rectification-confirm-birth-time",
|
||
"rectification-stop-and-review",
|
||
"rectification-close-case",
|
||
];
|
||
for (const name of toolNames) {
|
||
const tool = tools[name as keyof typeof tools] as unknown as { inputSchema: { safeParse(value: unknown): { success: boolean } } };
|
||
assert.ok(tool?.inputSchema, `${name} must expose an input schema`);
|
||
const malicious = {
|
||
caseId: CASE_ID,
|
||
userId: USER_ID,
|
||
birth_date: "1997-08-08",
|
||
latitude: 36.4,
|
||
longitude: 114.2,
|
||
timezone_offset: 8,
|
||
candidate_range: { start_time: "04:00", end_time: "06:00" },
|
||
events: [{ id: "e1", domain: "career", date: "2016-09" }],
|
||
scores: [10, 20],
|
||
confirmationAllowed: true,
|
||
};
|
||
const result = tool.inputSchema.safeParse(malicious);
|
||
assert.equal(result.success, false, `${name} must reject userId/birth/range/events/scores/permissions`);
|
||
}
|
||
});
|
||
|
||
type StreamChunk = {
|
||
type: string;
|
||
payload?: { toolName?: unknown; text?: unknown; args?: unknown; error?: unknown; result?: unknown };
|
||
};
|
||
|
||
type FakeStreamResult = {
|
||
fullStream: AsyncIterable<StreamChunk>;
|
||
totalUsage?: Promise<{ inputTokens?: number; outputTokens?: number }>;
|
||
};
|
||
|
||
function chunk(type: string, payload?: Record<string, unknown>): StreamChunk {
|
||
return { type, ...(payload ? { payload } : {}) };
|
||
}
|
||
|
||
function fakeAgentStream(chunks: StreamChunk[]) {
|
||
const streamResult: FakeStreamResult = {
|
||
fullStream: (async function* () {
|
||
for (const item of chunks) yield item;
|
||
})(),
|
||
totalUsage: Promise.resolve({ inputTokens: 10, outputTokens: 20 }),
|
||
};
|
||
return {
|
||
stream: async () => streamResult,
|
||
getSkill: async () => ({ name: RECTIFICATION_SKILL_NAME, instructions: "skill" }),
|
||
};
|
||
}
|
||
|
||
function runOptions(overrides: Partial<V9AgentRunOptions> = {}): {
|
||
options: V9AgentRunOptions;
|
||
emitted: Array<{ type: string }>;
|
||
billing: { reserved: number; completed: number; released: number };
|
||
} {
|
||
const emitted: Array<{ type: string }> = [];
|
||
const billing = { reserved: 0, completed: 0, released: 0 };
|
||
const accounting = fakeAccounting({
|
||
...receiptHandlers,
|
||
get_agentic_rectification_case_dossier: () => dossierFixture(),
|
||
append_agentic_rectification_turn: () => ({ turn_id: TURN_ID }),
|
||
finalize_agentic_rectification_turn: () => ({ turn_id: TURN_ID, status: "completed", idempotent: false }),
|
||
get_agentic_rectification_turn_receipt: () => null,
|
||
});
|
||
const optionsValue: V9AgentRunOptions = {
|
||
userId: USER_ID,
|
||
caseId: CASE_ID,
|
||
sessionId: SESSION_ID,
|
||
requestId: "aaaaaaaa-bbbb-4ccc-8ddd-eeeeeeeeeeee",
|
||
action: "evidence",
|
||
message: "2016年9月离开家去北京工作",
|
||
modelName: "gpt-4o-mini",
|
||
accounting: accounting.client,
|
||
billing: {
|
||
reserve: async () => { billing.reserved += 1; return { success: true, status: 200 }; },
|
||
complete: async () => { billing.completed += 1; return true; },
|
||
release: async () => { billing.released += 1; return true; },
|
||
},
|
||
emit: (event) => { emitted.push(event); },
|
||
buildAgent: async () => fakeAgentStream([]) as never,
|
||
...overrides,
|
||
};
|
||
return { options: optionsValue, emitted, billing };
|
||
}
|
||
|
||
test("agent receives the exact server-owned case id for tool calls", async () => {
|
||
let observedMessages: unknown[] = [];
|
||
const agent = fakeAgentStream([
|
||
chunk("start"),
|
||
chunk("tool-call", { toolName: "skill", args: { name: RECTIFICATION_SKILL_NAME } }),
|
||
chunk("tool-result", { toolName: "skill" }),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("text-delta", { text: "你好，我是生时校正助手。" }),
|
||
chunk("finish"),
|
||
]);
|
||
const { options } = runOptions({
|
||
action: "opening",
|
||
message: null,
|
||
buildAgent: async () => ({
|
||
...agent,
|
||
stream: async (messages: unknown[]) => {
|
||
observedMessages = messages;
|
||
return agent.stream();
|
||
},
|
||
}) as never,
|
||
});
|
||
|
||
const result = await runV9AgentTurn(options);
|
||
|
||
assert.equal(result.ok, true);
|
||
const openingPrompt = JSON.stringify(observedMessages);
|
||
assert.match(openingPrompt, new RegExp(CASE_ID));
|
||
assert.match(openingPrompt, /服务端 opening brief/);
|
||
// 旧：不要要求一次说完、不要举大学/工作/搬家的例子
|
||
// 新：窗口 + 六类领域 + 一条消息可以报多件
|
||
// 原因：BUG-604 决策 1，重算按轮计费
|
||
assert.match(openingPrompt, /一条消息可以报多件/);
|
||
assert.match(openingPrompt, new RegExp(OPENING_COLLECT_DOMAINS.join("、")));
|
||
assert.match(openingPrompt, /04:50–05:10|当前搜索窗口/);
|
||
assert.match(openingPrompt, /intake 声明的不确定档/);
|
||
assert.match(openingPrompt, /spokenPrompt 写出当前采集题/);
|
||
assert.match(openingPrompt, /不要提问/);
|
||
assert.doesNotMatch(openingPrompt, /接在正文末尾并写入聊天历史/);
|
||
assert.doesNotMatch(openingPrompt, /当前可询问范围/);
|
||
assert.doesNotMatch(openingPrompt, /说明你会通过已发生的人生事件来校正出生时间/);
|
||
});
|
||
|
||
test("server-loaded Skill is bound before the provider and the first model step only exposes read-case", async () => {
|
||
const skillInstructions = "immutable-skill-instructions-from-server";
|
||
let observedMessages: unknown[] = [];
|
||
let observedStreamOptions: {
|
||
prepareStep?: (input: { stepNumber: number }) => unknown;
|
||
modelSettings?: { maxOutputTokens?: number };
|
||
providerOptions?: Record<string, { thinking?: { type?: string } }>;
|
||
} = {};
|
||
const agent = fakeAgentStream([
|
||
chunk("start"),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("text-delta", { text: "你好，我先从一件你记得比较清楚的经历开始。" }),
|
||
chunk("finish"),
|
||
]);
|
||
const accounting = fakeAccounting({
|
||
...receiptHandlers,
|
||
get_agentic_rectification_case_dossier: () => dossierFixture({ turnCount: 0, turns: [] }),
|
||
append_agentic_rectification_turn: () => ({ turn_id: TURN_ID }),
|
||
finalize_agentic_rectification_turn: () => ({ turn_id: TURN_ID, status: "completed", idempotent: false }),
|
||
});
|
||
const { options, emitted, billing } = runOptions({
|
||
accounting: accounting.client,
|
||
buildAgent: async () => ({
|
||
...agent,
|
||
getSkill: async () => ({ name: RECTIFICATION_SKILL_NAME, instructions: skillInstructions }),
|
||
stream: async (messages: unknown[], streamOptions: typeof observedStreamOptions) => {
|
||
observedMessages = messages;
|
||
observedStreamOptions = streamOptions;
|
||
return agent.stream();
|
||
},
|
||
}) as never,
|
||
});
|
||
|
||
const result = await runV9AgentTurn(options);
|
||
|
||
assert.equal(result.ok, true);
|
||
assert.equal(result.skillLoaded, true);
|
||
assert.equal(result.errorCode, null);
|
||
assert.deepEqual(billing, { reserved: 1, completed: 1, released: 0 });
|
||
assert.match(JSON.stringify(observedMessages), /服务器已绑定当前 Case 的精确 Skill/);
|
||
assert.match(JSON.stringify(observedMessages), /不要重复调用 skill/);
|
||
assert.match(JSON.stringify(observedMessages), new RegExp(skillInstructions));
|
||
const firstStep = await observedStreamOptions.prepareStep?.({ stepNumber: 0 }) as {
|
||
activeTools?: unknown;
|
||
toolChoice?: unknown;
|
||
};
|
||
assert.deepEqual(firstStep, {
|
||
activeTools: ["rectification-read-case"],
|
||
toolChoice: "auto",
|
||
});
|
||
assert.equal(typeof firstStep?.toolChoice, "string");
|
||
assert.deepEqual(
|
||
(await observedStreamOptions.prepareStep?.({ stepNumber: 1 }) as { activeTools?: string[] }).activeTools?.includes("rectification-set-focus"),
|
||
true,
|
||
);
|
||
assert.ok(
|
||
((await observedStreamOptions.prepareStep?.({ stepNumber: 1 })) as { activeTools?: string[] }).activeTools?.includes("rectification-read-case"),
|
||
);
|
||
assert.equal(observedStreamOptions.modelSettings?.maxOutputTokens, 16_384);
|
||
assert.deepEqual(observedStreamOptions.providerOptions?.openai, { thinking: { type: "enabled" } });
|
||
assert.equal(emitted.filter((event) => event.type === "skill.bound").length, 1);
|
||
assert.equal(emitted.some((event) => event.type === "run.completed"), true);
|
||
assert.equal(
|
||
accounting.calls.filter((call) => call.fn === "insert_agentic_rectification_skill_run_receipt").length,
|
||
1,
|
||
);
|
||
});
|
||
|
||
test("first turn with a bound immutable Skill completes and persists receipts", async () => {
|
||
const accounting = fakeAccounting({
|
||
...receiptHandlers,
|
||
get_agentic_rectification_case_dossier: () => dossierFixture({ turnCount: 0 }),
|
||
append_agentic_rectification_turn: () => ({ turn_id: TURN_ID }),
|
||
finalize_agentic_rectification_turn: () => ({ turn_id: TURN_ID, status: "completed", idempotent: false }),
|
||
});
|
||
const { options, emitted, billing } = runOptions({
|
||
accounting: accounting.client,
|
||
buildAgent: async () => fakeAgentStream([
|
||
chunk("start"),
|
||
chunk("tool-call", { toolName: "skill", args: { name: RECTIFICATION_SKILL_NAME } }),
|
||
chunk("tool-result", { toolName: "skill" }),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("text-delta", { text: "你好，我是生时校正助手。" }),
|
||
chunk("finish"),
|
||
]) as never,
|
||
});
|
||
const result = await runV9AgentTurn(options);
|
||
assert.equal(result.ok, true);
|
||
assert.equal(result.turnStatus, "completed");
|
||
assert.equal(result.skillLoaded, true);
|
||
assert.equal(billing.completed, 1);
|
||
assert.equal(billing.released, 0);
|
||
const types = emitted.map((event) => event.type);
|
||
assert.ok(types.includes("run.started"));
|
||
assert.ok(types.includes("skill.bound"));
|
||
assert.ok(types.includes("case.loaded"));
|
||
assert.ok(types.includes("intent.classified"));
|
||
assert.ok(types.includes("answer.composed"));
|
||
assert.ok(types.includes("billing.settled"));
|
||
assert.ok(types.includes("answer.delta"));
|
||
assert.ok(types.includes("run.completed"));
|
||
// Reasoning and raw chunks are dropped.
|
||
const accountingCalls = accounting.calls.map((call) => call.fn);
|
||
assert.ok(accountingCalls.includes("insert_agentic_rectification_run_phase"));
|
||
assert.ok(accountingCalls.includes("finalize_agentic_rectification_turn"));
|
||
});
|
||
|
||
test("collect_spoken stem is persisted on the same turn via asked_turn_id, not assistant_message", async () => {
|
||
const greeting = "你好，我是生时校正助手。";
|
||
const collectDossier = dossierFixture({
|
||
conversationSummary: conversationSummaryFixture({
|
||
activeFocus: activeFocusFixture({
|
||
intent: "collect_method_evidence",
|
||
questionId: "collect:unknown:collect_method_evidence",
|
||
expectedAnswerSchema: {
|
||
prompt: GENERIC_COLLECT_QUESTION,
|
||
collect: true,
|
||
},
|
||
}),
|
||
}),
|
||
});
|
||
const accounting = fakeAccounting({
|
||
...receiptHandlers,
|
||
get_agentic_rectification_case_dossier: () => collectDossier,
|
||
append_agentic_rectification_turn: () => ({ turn_id: TURN_ID }),
|
||
finalize_agentic_rectification_turn: () => ({ turn_id: TURN_ID, status: "completed", idempotent: false }),
|
||
});
|
||
const { options, emitted } = runOptions({
|
||
action: "opening",
|
||
message: null,
|
||
accounting: accounting.client,
|
||
buildAgent: async () => fakeAgentStream([
|
||
chunk("start"),
|
||
chunk("tool-call", { toolName: "skill", args: { name: RECTIFICATION_SKILL_NAME } }),
|
||
chunk("tool-result", { toolName: "skill" }),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("text-delta", { text: greeting }),
|
||
chunk("finish"),
|
||
]) as never,
|
||
});
|
||
const result = await runV9AgentTurn(options);
|
||
assert.equal(result.ok, true);
|
||
// 旧：模型一句招呼原样落库 → 新：开场正文必须过六类领域验收，不合格则换成模板
|
||
// 原因：BUG-604 开场 ≤4 句、≥5 类、不含年份
|
||
const spoken = openingSpokenBody(["04:50", "05:10"]);
|
||
assert.equal(isAcceptableOpeningBody(result.answerText), true);
|
||
assert.equal(result.answerText, spoken);
|
||
const replaced = emitted.find((event) => (
|
||
event.type === "answer.delta" && (event as { replace?: unknown }).replace === true
|
||
)) as { text?: string } | undefined;
|
||
assert.notEqual(replaced?.text, composeCollectSpokenAssistantText(greeting, GENERIC_COLLECT_QUESTION));
|
||
const finalize = accounting.calls.find((call) => call.fn === "finalize_agentic_rectification_turn");
|
||
assert.equal(finalize?.args.p_assistant_message, spoken);
|
||
assert.equal(
|
||
accounting.calls.some((call) => (
|
||
call.fn === "set_agentic_rectification_conversation_focus"
|
||
&& call.args.p_asked_turn_id === TURN_ID
|
||
)),
|
||
true,
|
||
);
|
||
});
|
||
|
||
test("framework getSkill failure is a controlled retry and then a failed turn", async () => {
|
||
const { options, billing } = runOptions({
|
||
buildAgent: async () => ({
|
||
stream: async () => ({ fullStream: (async function* () {})() }),
|
||
getSkill: async () => null,
|
||
}) as never,
|
||
});
|
||
const result = await runV9AgentTurn(options);
|
||
assert.equal(result.ok, false);
|
||
assert.equal(result.errorCode, "skill_not_loaded");
|
||
assert.equal(billing.released, 1);
|
||
});
|
||
|
||
test("distinct evidence calls in one natural turn are not mistaken for a repeated tool loop", async () => {
|
||
const firstEvidenceId = EVIDENCE_ID;
|
||
const secondEvidenceId = "88888888-8888-4888-8888-888888888888";
|
||
const { options, billing } = runOptions({
|
||
buildAgent: async () => fakeAgentStream([
|
||
chunk("start"),
|
||
chunk("tool-call", { toolName: "skill", args: { name: RECTIFICATION_SKILL_NAME } }),
|
||
chunk("tool-result", { toolName: "skill" }),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("tool-call", {
|
||
toolName: "rectification-propose-evidence",
|
||
args: { caseId: CASE_ID, quote: "2016年9月离开家去北京开始工作", proposedKind: "career_entry" },
|
||
}),
|
||
chunk("tool-result", { toolName: "rectification-propose-evidence" }),
|
||
chunk("tool-call", { toolName: "rectification-confirm-evidence", args: { caseId: CASE_ID, evidenceId: firstEvidenceId } }),
|
||
chunk("tool-result", { toolName: "rectification-confirm-evidence" }),
|
||
chunk("tool-call", {
|
||
toolName: "rectification-propose-evidence",
|
||
args: { caseId: CASE_ID, quote: "2020年又搬到了上海", proposedKind: "relocation" },
|
||
}),
|
||
chunk("tool-result", { toolName: "rectification-propose-evidence" }),
|
||
chunk("tool-call", { toolName: "rectification-confirm-evidence", args: { caseId: CASE_ID, evidenceId: secondEvidenceId } }),
|
||
chunk("tool-result", { toolName: "rectification-confirm-evidence" }),
|
||
chunk("text-delta", { text: "这两件事已经分别纳入当前校正。" }),
|
||
chunk("finish"),
|
||
]) as never,
|
||
});
|
||
const result = await runV9AgentTurn(options);
|
||
assert.equal(result.ok, true);
|
||
assert.equal(result.errorCode, null);
|
||
assert.equal(billing.completed, 1);
|
||
assert.deepEqual(result.toolsUsed.filter((name) => name.includes("evidence")), [
|
||
"rectification-propose-evidence",
|
||
"rectification-confirm-evidence",
|
||
]);
|
||
});
|
||
|
||
test("a repeated identical tool call is treated as idempotent and does not abort", async () => {
|
||
const { options, emitted, billing } = runOptions({
|
||
buildAgent: async () => fakeAgentStream([
|
||
chunk("start"),
|
||
chunk("tool-call", { toolName: "skill", args: { name: RECTIFICATION_SKILL_NAME } }),
|
||
chunk("tool-result", { toolName: "skill" }),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("text-delta", { text: "已经记下。" }),
|
||
chunk("finish"),
|
||
]) as never,
|
||
});
|
||
const result = await runV9AgentTurn(options);
|
||
assert.equal(result.ok, true);
|
||
assert.equal(result.errorCode, null);
|
||
assert.match(result.answerText, /已经记下|请继续说下一件/);
|
||
assert.equal(billing.completed, 1);
|
||
assert.equal(billing.released, 0);
|
||
assert.equal(emitted.some((event) => event.type === "attempt.reset"), false);
|
||
assert.equal(emitted.some((event) => event.type === "run.failed"), false);
|
||
assert.equal(
|
||
emitted.filter((event) => event.type === "tool.activity"
|
||
&& (event as { tool?: string; status?: string }).tool === "rectification-read-case"
|
||
&& (event as { tool?: string; status?: string }).status === "started").length,
|
||
1,
|
||
);
|
||
});
|
||
|
||
test("a failed opening does not let the next turn skip the server Skill load gate", async () => {
|
||
let streamCount = 0;
|
||
const accounting = fakeAccounting({
|
||
...receiptHandlers,
|
||
get_agentic_rectification_case_dossier: () => dossierFixture({
|
||
turnCount: 1,
|
||
turns: [{
|
||
id: TURN_ID,
|
||
role: "user",
|
||
text: "（开场）",
|
||
status: "failed",
|
||
created_at: "2026-08-12T10:00:00.000Z",
|
||
completed_at: null,
|
||
}],
|
||
}),
|
||
append_agentic_rectification_turn: () => ({ turn_id: TURN_ID }),
|
||
finalize_agentic_rectification_turn: () => ({ turn_id: TURN_ID, status: "failed", idempotent: false }),
|
||
});
|
||
const { options, billing } = runOptions({
|
||
accounting: accounting.client,
|
||
buildAgent: async () => ({
|
||
getSkill: async () => null,
|
||
stream: async () => {
|
||
streamCount += 1;
|
||
return { fullStream: (async function* () {})() };
|
||
},
|
||
}) as never,
|
||
});
|
||
|
||
const result = await runV9AgentTurn(options);
|
||
|
||
assert.equal(result.ok, false);
|
||
assert.equal(result.errorCode, "skill_not_loaded");
|
||
assert.equal(streamCount, 0);
|
||
assert.equal(billing.released, 1);
|
||
});
|
||
|
||
test("same evidence + range fingerprints reuse the cached candidate snapshot", async () => {
|
||
const inputReceipt = {
|
||
receipt_version: "candidate-decision-receipt-v2",
|
||
policy_version: "rectification-candidate-policy-v2",
|
||
selection_allowed: true,
|
||
acceptance_allowed: true,
|
||
confirmation_allowed: false,
|
||
representative_candidate_id: null,
|
||
overall_confidence: "medium",
|
||
inference_state: { revision: 1, candidates: [{ id: "05:02", posterior_score: 10 }] },
|
||
};
|
||
const composedReceipt = {
|
||
...inputReceipt,
|
||
inference_state: { revision: 2, candidates: [{ id: "05:02", posterior_score: 18 }] },
|
||
decision_state_fingerprint: "a".repeat(64),
|
||
};
|
||
const accounting = fakeAccounting({
|
||
persist_agentic_rectification_candidate_v2: () => ({
|
||
...candidateSnapshotFixture(),
|
||
cached: true,
|
||
decision_receipt: composedReceipt,
|
||
}),
|
||
});
|
||
const cached = await persistV9Candidate(accounting.client, USER_ID, CASE_ID, {
|
||
engineResultId: "engine-1",
|
||
algorithmVersion: "rectification-v5",
|
||
evidenceFingerprint: "b".repeat(64),
|
||
rangeFingerprint: "c".repeat(64),
|
||
skillVersion: "9.0.0",
|
||
candidateRange: CANDIDATE_RANGE,
|
||
eventContractVersion: "rectification-event-contract-v2",
|
||
policyVersion: "rectification-candidate-policy-v2",
|
||
candidates: [{ candidateId: CANDIDATE_ID, rank: 1, time: "05:02", relativeSupport: 58, tiedMinuteCount: 2 }],
|
||
decisionReceipt: inputReceipt,
|
||
executionLedger: [{ method: "d1-rashi", status: "executed" }],
|
||
});
|
||
assert.equal(cached.cached, true);
|
||
assert.equal(cached.resultId, RESULT_ID);
|
||
const call = accounting.calls.find((item) => item.fn === "persist_agentic_rectification_candidate_v2");
|
||
assert.ok(call);
|
||
assert.equal(call.args.p_evidence_ledger_fingerprint, "b".repeat(64));
|
||
assert.equal(call.args.p_skill_version, "9.0.0");
|
||
assert.equal(call.args.p_event_contract_version, "rectification-event-contract-v2");
|
||
assert.equal(call.args.p_decision_policy_version, "rectification-candidate-policy-v2");
|
||
assert.deepEqual(call.args.p_decision_receipt, inputReceipt);
|
||
assert.notDeepEqual(cached.decisionReceipt, inputReceipt);
|
||
assert.deepEqual(cached.decisionReceipt, composedReceipt);
|
||
assert.equal(
|
||
(cached.decisionReceipt.inference_state as { revision?: number }).revision,
|
||
2,
|
||
);
|
||
assert.equal(cached.decisionReceipt.decision_state_fingerprint, "a".repeat(64));
|
||
assert.deepEqual(call.args.p_execution_ledger, cached.executionLedger);
|
||
assert.equal("p_selection_allowed" in call.args, false);
|
||
assert.equal("p_confirmation_allowed" in call.args, false);
|
||
assert.equal("p_representative_time" in call.args, false);
|
||
assert.equal("p_margin_percent" in call.args, false);
|
||
assert.equal("p_decision_state_fingerprint" in call.args, false);
|
||
});
|
||
|
||
test("accept-candidate requires a server-persisted result; no tool means no minute", async () => {
|
||
const accounting = fakeAccounting({
|
||
...receiptHandlers,
|
||
get_agentic_rectification_case_dossier: () => dossierFixture({ latestResult: null }),
|
||
accept_agentic_rectification_candidate_for_case_v2: () => {
|
||
throw new Error("agentic_rectification_candidate_not_found");
|
||
},
|
||
});
|
||
const tools = createRectificationV9Tools({
|
||
userId: USER_ID,
|
||
caseId: CASE_ID,
|
||
turnId: TURN_ID,
|
||
accounting: accounting.client as never,
|
||
});
|
||
await assert.rejects(
|
||
(tools["rectification-accept-candidate"] as unknown as { execute(input: unknown): Promise<unknown> }).execute({
|
||
caseId: CASE_ID,
|
||
resultId: RESULT_ID,
|
||
candidateId: CANDIDATE_ID,
|
||
}),
|
||
// 原值 candidate_not_found；新值 result_identity_read_only；原因：无来源在写 RPC 前即拒绝，仍不得产生候选。
|
||
(error: unknown) => error instanceof Error && error.message.includes("result_identity_read_only"),
|
||
);
|
||
});
|
||
|
||
test("new evidence lets the agent choose diagnostics/compare tools autonomously", () => {
|
||
// The tool layer exposes read-diagnostics and compare-candidates; there is
|
||
// no hard-coded orchestration forcing a scan before score.
|
||
const tools = createRectificationV9Tools({
|
||
userId: USER_ID,
|
||
caseId: CASE_ID,
|
||
turnId: TURN_ID,
|
||
accounting: fakeAccounting({}).client as never,
|
||
});
|
||
assert.ok("rectification-compare-candidates" in tools);
|
||
assert.ok("rectification-read-diagnostics" in tools);
|
||
assert.ok("rectification-offer-candidates" in tools);
|
||
assert.ok(!("rectification-scan" in tools));
|
||
assert.ok(!("rectification-gate" in tools));
|
||
});
|
||
|
||
test("agent construction wires the pinned skill and the ten v9 tools", () => {
|
||
const model = {
|
||
id: "test-model",
|
||
label: "Test",
|
||
description: "",
|
||
creditCost: 1,
|
||
isDefault: true,
|
||
mode: "compatible" as const,
|
||
model: { provider: "openai", name: "gpt-4o-mini", modelId: "gpt-4o-mini" } as never,
|
||
};
|
||
const accounting = fakeAccounting({});
|
||
const agent = getRectificationV9Agent(model, {
|
||
userId: USER_ID,
|
||
caseId: CASE_ID,
|
||
turnId: TURN_ID,
|
||
accounting: accounting.client as never,
|
||
});
|
||
assert.equal(agent.id, "rectification-v9-test-model");
|
||
assert.ok(agent);
|
||
});
|
||
|
||
test("rectification has no reply-regeneration agent or read-only tool set (BUG-1056)", () => {
|
||
// 原值: 「reply regeneration is a separate Jyotisha agent with only read-case access」——
|
||
// 断言 createRectificationV9ReadOnlyTools 只含 read-case、getRectificationV9RegenerationAgent
|
||
// 的 id 与提示词。
|
||
// 新值: 两个导出都不存在，提示词里也没有重新生成的指令。
|
||
// 原因: TASK-rectification-grounding-20260927 P1 产品决定去掉生时校正的「重试（重新生成）」；
|
||
// 这两个导出只服务已删除的重新生成接口（BUG-1056：盲写、不校验就落库）。
|
||
assert.equal("getRectificationV9RegenerationAgent" in agenticRectification, false);
|
||
assert.equal("createRectificationV9ReadOnlyTools" in rectificationV9Tools, false);
|
||
assert.doesNotMatch(agentSource, /这不是新一轮校正/);
|
||
assert.doesNotMatch(agentSource, /重新生成最近一条 Agent 正文/);
|
||
});
|
||
|
||
test("set-focus after spoken text does not append a second paragraph", async () => {
|
||
// 原值: first 第二句「范围收到 05:00 到 05:10。」。
|
||
// 新值: 第二句「这件事拿去和星盘对照了。」。
|
||
// 原因: BUG-1055 / P3 起证据轮模型正文里与当轮服务端事实不符的时刻句整句不落库；本夹具无
|
||
// 候选快照，05:00–05:10 是手写数字。本用例锁的是 set-focus 之后不接第二段，与数字无关。
|
||
const first = "2016 年 9 月去北京工作，记下了。这件事拿去和星盘对照了。";
|
||
const second = "2016 年那件事对校正很有帮助，我们再对一下搬家。";
|
||
const accounting = fakeAccounting({
|
||
...receiptHandlers,
|
||
get_agentic_rectification_case_dossier: () => dossierFixture({ turnCount: 0 }),
|
||
append_agentic_rectification_turn: () => ({ turn_id: TURN_ID }),
|
||
finalize_agentic_rectification_turn: () => ({ turn_id: TURN_ID, status: "completed", idempotent: false }),
|
||
});
|
||
const { options, emitted } = runOptions({
|
||
accounting: accounting.client,
|
||
buildAgent: async () => fakeAgentStream([
|
||
chunk("start"),
|
||
chunk("tool-call", { toolName: "skill", args: { name: RECTIFICATION_SKILL_NAME } }),
|
||
chunk("tool-result", { toolName: "skill" }),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("text-delta", { text: first }),
|
||
chunk("tool-call", { toolName: "rectification-set-focus" }),
|
||
chunk("tool-result", { toolName: "rectification-set-focus" }),
|
||
chunk("text-delta", { text: second }),
|
||
chunk("finish"),
|
||
]) as never,
|
||
});
|
||
const result = await runV9AgentTurn(options);
|
||
assert.equal(result.ok, true);
|
||
assert.equal(result.answerText, first);
|
||
assert.equal(result.answerText.includes(second), false);
|
||
const deltas = emitted.filter((event) => event.type === "answer.delta") as Array<{
|
||
text?: string;
|
||
replace?: boolean;
|
||
}>;
|
||
assert.equal(deltas.some((item) => (item.text ?? "").includes(second)), false);
|
||
assert.equal(deltas.some((item) => item.replace === true && item.text === ""), false);
|
||
});
|
||
|
||
test("agent-run strips question sentences when this turn owns the focus", async () => {
|
||
// 原值: body 首句「范围已经收到，收在 05:00–05:10。」，断言落库与 replace 都等于该句。
|
||
// 新值: 首句「2016 年 9 月去北京工作，这件事拿去和星盘对照。」，断言同样等于首句。
|
||
// 原因: BUG-1055 / P3：本夹具没有候选快照，05:00–05:10 不是服务端事实，整句会被数字白名单
|
||
// 丢弃。本用例锁的是「本轮拥有焦点时剪掉问句」，换成不带时刻的首句后剪问句断言不变。
|
||
const body = "2016 年 9 月去北京工作，这件事拿去和星盘对照。你大概哪一年搬过家？有年份就行。";
|
||
const stem = "你大概是哪一年搬的家？";
|
||
const collectDossier = dossierFixture({
|
||
conversationSummary: conversationSummaryFixture({
|
||
activeFocus: activeFocusFixture({
|
||
askedTurnId: TURN_ID,
|
||
intent: "collect_method_evidence",
|
||
expectedAnswerSchema: { prompt: stem, collect: true },
|
||
}),
|
||
}),
|
||
});
|
||
const accounting = fakeAccounting({
|
||
...receiptHandlers,
|
||
get_agentic_rectification_case_dossier: () => collectDossier,
|
||
append_agentic_rectification_turn: () => ({ turn_id: TURN_ID }),
|
||
finalize_agentic_rectification_turn: () => ({ turn_id: TURN_ID, status: "completed", idempotent: false }),
|
||
});
|
||
const { options, emitted } = runOptions({
|
||
accounting: accounting.client,
|
||
buildAgent: async () => fakeAgentStream([
|
||
chunk("start"),
|
||
chunk("tool-call", { toolName: "skill", args: { name: RECTIFICATION_SKILL_NAME } }),
|
||
chunk("tool-result", { toolName: "skill" }),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("text-delta", { text: body }),
|
||
chunk("finish"),
|
||
]) as never,
|
||
});
|
||
const result = await runV9AgentTurn(options);
|
||
assert.equal(result.ok, true);
|
||
assert.equal(result.answerText, "2016 年 9 月去北京工作，这件事拿去和星盘对照。");
|
||
const replaced = emitted.find((event) => (
|
||
event.type === "answer.delta" && (event as { replace?: unknown }).replace === true
|
||
)) as { text?: string } | undefined;
|
||
assert.equal(replaced?.text, "2016 年 9 月去北京工作，这件事拿去和星盘对照。");
|
||
});
|
||
|
||
test("agent-run uses collectHandoff when the owned-focus body is only a question", async () => {
|
||
const stem = "你大概是哪一年搬的家？";
|
||
const collectDossier = dossierFixture({
|
||
conversationSummary: conversationSummaryFixture({
|
||
activeFocus: activeFocusFixture({
|
||
askedTurnId: TURN_ID,
|
||
intent: "collect_method_evidence",
|
||
expectedAnswerSchema: { prompt: stem, collect: true },
|
||
}),
|
||
}),
|
||
});
|
||
const accounting = fakeAccounting({
|
||
...receiptHandlers,
|
||
get_agentic_rectification_case_dossier: () => collectDossier,
|
||
append_agentic_rectification_turn: () => ({ turn_id: TURN_ID }),
|
||
finalize_agentic_rectification_turn: () => ({ turn_id: TURN_ID, status: "completed", idempotent: false }),
|
||
});
|
||
const { options } = runOptions({
|
||
accounting: accounting.client,
|
||
buildAgent: async () => fakeAgentStream([
|
||
chunk("start"),
|
||
chunk("tool-call", { toolName: "skill", args: { name: RECTIFICATION_SKILL_NAME } }),
|
||
chunk("tool-result", { toolName: "skill" }),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("text-delta", { text: "你大概哪一年搬过家？" }),
|
||
chunk("finish"),
|
||
]) as never,
|
||
});
|
||
const result = await runV9AgentTurn(options);
|
||
assert.equal(result.ok, true);
|
||
assert.equal(result.answerText, RECTIFICATION_USER_COPY.collectHandoff);
|
||
});
|
||
|
||
test("buildOpeningBrief names the search window, intake source, and six domains", () => {
|
||
const dossier = parseV9CaseDossier(dossierFixture({
|
||
candidateRange: { start_time: "04:45", end_time: "05:15" },
|
||
evidence: [],
|
||
evidenceCount: 0,
|
||
}));
|
||
assert.ok(dossier);
|
||
const brief = buildOpeningBrief(dossier);
|
||
assert.match(brief, /当前搜索窗口：04:45–05:15/);
|
||
assert.match(brief, /intake 声明的不确定档/);
|
||
for (const domain of OPENING_COLLECT_DOMAINS) {
|
||
assert.match(brief, new RegExp(domain));
|
||
}
|
||
// 原值: /先说你最容易想起的一两件，年月大概就行/（brief 让模型照抄旧题干）
|
||
// 新值: brief 说明开场题干由服务端固定，正文两句大白话、不用行话、不重复例子
|
||
// 原因: BUG-1049 开场题干改为服务端固定（带例子与示例回答），模型只写正文；brief 仍不得含年份
|
||
assert.match(brief, /开场题干由服务端固定/);
|
||
assert.match(brief, /我们来把你的出生时间缩小到更准的范围，现在先在 04:45–05:15 之间找/);
|
||
assert.match(brief, /做法很简单：你说几件人生里的大事和大概年月，我拿去和星盘对照/);
|
||
assert.match(brief, /正文不要重复这些例子/);
|
||
assert.doesNotMatch(brief, /最后给区间和代表分钟/);
|
||
assert.doesNotMatch(brief, /先说你最容易想起的一两件/);
|
||
assert.doesNotMatch(brief, /不要要求一次说完/);
|
||
assert.doesNotMatch(brief, /不要举大学、工作、搬家的例子/);
|
||
assert.doesNotMatch(brief, /(?:19|20)\d{2}/);
|
||
});
|
||
|
||
test("evidence persist keeps the first sentence when the model writes three", async () => {
|
||
const accounting = fakeAccounting({
|
||
...receiptHandlers,
|
||
get_agentic_rectification_case_dossier: () => dossierFixture({
|
||
conversationSummary: conversationSummaryFixture({
|
||
activeFocus: activeFocusFixture({ intent: "collect_method_evidence" }),
|
||
}),
|
||
}),
|
||
append_agentic_rectification_turn: () => ({ turn_id: TURN_ID }),
|
||
finalize_agentic_rectification_turn: () => ({ turn_id: TURN_ID, status: "completed", idempotent: false }),
|
||
});
|
||
// 原夹具：read-case 后直接写三句「记下了」，无 batch。
|
||
// 新夹具：batch completed 后再写那三句。
|
||
// 原因：BUG-635 把无写入的「记下了」收口成「这件我还没记上」；本用例锁的是已 persist 后只留第一句。
|
||
const { options } = runOptions({
|
||
action: "evidence",
|
||
accounting: accounting.client,
|
||
buildAgent: async () => fakeAgentStream([
|
||
chunk("start"),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("tool-call", { toolName: "rectification-record-evidence-batch", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", {
|
||
toolName: "rectification-record-evidence-batch",
|
||
result: {
|
||
accepted_recaps: [
|
||
{ display_date_label: "2016 年 9 月", event_phrase: "入学" },
|
||
{ display_date_label: "2020 年 6 月", event_phrase: "毕业" },
|
||
],
|
||
},
|
||
}),
|
||
chunk("tool-call", { toolName: "rectification-set-focus", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-set-focus" }),
|
||
chunk("text-delta", {
|
||
text: "记下了：2016 年 9 月入学、2020 年 6 月毕业。这对校正很有帮助。接下来我们继续。",
|
||
}),
|
||
chunk("finish"),
|
||
]) as never,
|
||
});
|
||
const result = await runV9AgentTurn(options);
|
||
assert.equal(result.ok, true);
|
||
assert.match(result.answerText, /^记下了：2016 年 9 月入学、2020 年 6 月毕业。/);
|
||
assert.doesNotMatch(result.answerText, /很有帮助|很有价值|很有分量|特别有用/);
|
||
assert.doesNotMatch(result.answerText, /接下来我们继续/);
|
||
const finalize = accounting.calls.find((call) => call.fn === "finalize_agentic_rectification_turn");
|
||
assert.match(String(finalize?.args.p_assistant_message ?? ""), /^记下了：2016 年 9 月入学、2020 年 6 月毕业。/);
|
||
assert.doesNotMatch(String(finalize?.args.p_assistant_message ?? ""), /很有帮助|特别有用/);
|
||
});
|
||
|
||
test("delivery persist keeps three sentences when the model writes four", async () => {
|
||
const accounting = fakeAccounting({
|
||
...receiptHandlers,
|
||
get_agentic_rectification_case_dossier: () => dossierFixture({
|
||
conversationSummary: conversationSummaryFixture({
|
||
declinedSkippedTopics: [
|
||
"family",
|
||
"education",
|
||
"finance",
|
||
"relocation",
|
||
"health_pressure",
|
||
"career",
|
||
"relationship",
|
||
"occupation",
|
||
].map((domain) => ({
|
||
target_domain: domain,
|
||
status: "declined",
|
||
intent: "collect_method_evidence",
|
||
})),
|
||
}),
|
||
latestResult: candidateSnapshotFixture({
|
||
selectionAllowed: true,
|
||
representativeTime: "05:02",
|
||
decisionReceipt: {
|
||
session_outcome: "adopt_representative",
|
||
next_action: "ready_to_adopt",
|
||
stop_reason: "probe_pool_exhausted",
|
||
can_adopt: true,
|
||
// 原值: 无 event_fit_rate。新值: 8/10、80%。原因: BUG-1055 / P3 数字白名单只放行当轮
|
||
// 服务端事实；交付句里的「吻合率 80%」需要回执里真有这个吻合率才落库。
|
||
event_fit_rate: { matched: 8, total: 10, percent: 80, band: "high" },
|
||
},
|
||
}),
|
||
evidence: [
|
||
{
|
||
id: "e-career",
|
||
source_turn_id: TURN_ID,
|
||
subject: "self",
|
||
event_kind: "career_entry",
|
||
domain: "career",
|
||
occurred_from: "2016-09-01",
|
||
occurred_to: null,
|
||
date_precision: "month",
|
||
summary: "2016年9月入职",
|
||
status: "confirmed",
|
||
supersedes_evidence_id: null,
|
||
created_at: "2026-08-12T10:00:06.000Z",
|
||
},
|
||
{
|
||
id: "e-edu",
|
||
source_turn_id: TURN_ID,
|
||
subject: "self",
|
||
event_kind: "education_start",
|
||
domain: "education",
|
||
occurred_from: "2014-09-01",
|
||
occurred_to: null,
|
||
date_precision: "month",
|
||
summary: "2014年9月入学",
|
||
status: "confirmed",
|
||
supersedes_evidence_id: null,
|
||
created_at: "2026-08-12T10:00:06.000Z",
|
||
},
|
||
{
|
||
id: "e-move",
|
||
source_turn_id: TURN_ID,
|
||
subject: "self",
|
||
event_kind: "home_change",
|
||
domain: "relocation",
|
||
occurred_from: "2020-06-01",
|
||
occurred_to: null,
|
||
date_precision: "month",
|
||
summary: "2020年6月搬家",
|
||
status: "confirmed",
|
||
supersedes_evidence_id: null,
|
||
created_at: "2026-08-12T10:00:06.000Z",
|
||
},
|
||
],
|
||
}),
|
||
get_agentic_rectification_case_compute: () => computeFixture(),
|
||
append_agentic_rectification_turn: () => ({ turn_id: TURN_ID }),
|
||
finalize_agentic_rectification_turn: () => ({ turn_id: TURN_ID, status: "completed", idempotent: false }),
|
||
});
|
||
// 原值: 首句「目前范围 04:49–04:53。」。新值: 「代表分钟是 05:02。」。原因: BUG-1055 / P3，
|
||
// 本夹具的服务端范围与代表分钟都是 05:02，04:49–04:53 是手写数字，会整句不落库。
|
||
const four = [
|
||
"代表分钟是 05:02。",
|
||
"对照了 8 件经历，事件吻合率 80%。",
|
||
"这只是代表性候选，不是已确认的唯一出生分钟。",
|
||
"还可以再看一眼分盘。",
|
||
].join("");
|
||
const { options } = runOptions({
|
||
action: "evidence",
|
||
accounting: accounting.client,
|
||
buildAgent: async () => fakeAgentStream([
|
||
chunk("start"),
|
||
chunk("tool-call", { toolName: "rectification-read-case", args: { caseId: CASE_ID } }),
|
||
chunk("tool-result", { toolName: "rectification-read-case" }),
|
||
chunk("text-delta", { text: four }),
|
||
chunk("finish"),
|
||
]) as never,
|
||
});
|
||
const result = await runV9AgentTurn(options);
|
||
assert.equal(result.ok, true);
|
||
const sentences = result.answerText.split(/(?<=。)/).filter(Boolean);
|
||
assert.equal(sentences.length, 3);
|
||
assert.doesNotMatch(result.answerText, /还可以再看一眼分盘/);
|
||
const finalize = accounting.calls.find((call) => call.fn === "finalize_agentic_rectification_turn");
|
||
assert.equal(String(finalize?.args.p_assistant_message ?? "").split(/(?<=。)/).filter(Boolean).length, 3);
|
||
});
|