Jesse_Chen
3198fb6b2b
Settled messages no longer rebuild on every token, Inter is actually requested, and admin routes drop the 33 KB chat stylesheet. Root force-dynamic is gone so public shells can prerender without changing the no-store Cache-Control contract. Co-authored-by: Cursor <cursoragent@cursor.com>
976 lines
37 KiB
TypeScript
976 lines
37 KiB
TypeScript
import assert from "node:assert/strict";
|
|
import { readFileSync } from "node:fs";
|
|
import test from "node:test";
|
|
|
|
import { applyRectificationChoice } from "../src/lib/rectification-agentic/v9/answer-choice.ts";
|
|
import { choiceCardFromCaseDossier } from "../src/lib/rectification-agentic/v9/interview-state.ts";
|
|
import {
|
|
CHOICE_ACTION,
|
|
STOP_ACTION,
|
|
composeChoiceNarration,
|
|
isStructuredChoiceUserText,
|
|
quoteIsFromAssistantQuestion,
|
|
shouldContinueAfterStructuredChoice,
|
|
userVisibleChoiceLine,
|
|
} from "../src/lib/rectification-agentic/v9/choice-action.ts";
|
|
import { isNearBottom, shouldFollowLatest, shouldShowJumpToLatest } from "../src/lib/rectification-sticky-scroll.ts";
|
|
import { projectTurnDecision, TURN_DECISION_MAX_BYTES, turnDecisionByteLength } from "../src/lib/rectification-agentic/v9/turn-decision.ts";
|
|
import {
|
|
mapModelFinishToErrorCode,
|
|
userFacingRunFailure,
|
|
isIncompleteRunBanner,
|
|
} from "../src/lib/rectification-agentic/v9/run-diagnostic.ts";
|
|
import { buildInferenceState } from "../src/lib/rectification-agentic/core/build-state.ts";
|
|
import { parseV9CaseDossier, RectificationToolServiceError } from "../src/lib/rectification-agentic/v9/tool-service.ts";
|
|
import {
|
|
CASE_ID,
|
|
EVIDENCE_ID,
|
|
FOCUS_ID,
|
|
SESSION_ID,
|
|
TURN_ID,
|
|
USER_ID,
|
|
activeFocusFixture,
|
|
candidateSnapshotFixture,
|
|
computeFixture,
|
|
conversationSummaryFixture,
|
|
dossierFixture,
|
|
fakeAccounting,
|
|
receiptHandlers,
|
|
} from "./rectification-v9-test-support.ts";
|
|
|
|
const ACTION_ID = "aaaaaaaa-aaaa-4aaa-8aaa-aaaaaaaaaaaa";
|
|
const QUESTION_ID = "question-1";
|
|
const NEXT_FOCUS_ID = "acacacac-acac-4cac-8cac-acacacacacac";
|
|
const EDUCATION_ID = "44444444-4444-4444-8444-444444444445";
|
|
const RELATIONSHIP_ID = "44444444-4444-4444-8444-444444444446";
|
|
const CAREER_EXIT_ID = "44444444-4444-4444-8444-444444444447";
|
|
const RELATIONSHIP_END_ID = "44444444-4444-4444-8444-444444444448";
|
|
|
|
const STYLE_OPTIONS = [
|
|
{ label: "明确发生且时间吻合", answer_class: "yes" as const },
|
|
{ label: "发生过但程度较弱", answer_class: "weak_yes" as const },
|
|
{ label: "明确没有发生", answer_class: "no" as const },
|
|
{ label: "这段记不清楚", answer_class: "unsure" as const },
|
|
];
|
|
|
|
const RELOCATION_2015_PROBE = {
|
|
id: "p-reloc-2015",
|
|
semantic_key: "relocation.2015.05.dasha_boundary",
|
|
candidate_split_hash: "reloc-2015-split",
|
|
domain: "relocation",
|
|
year: 2015,
|
|
month: 5,
|
|
question: "2015 年 5 月前后有没有搬家或长期住到外地？",
|
|
candidate_ids: ["05:00", "05:10"],
|
|
expected_outcomes: [
|
|
{ answer_class: "yes" as const, supports: ["05:00"], conflicts: ["05:10"] },
|
|
{ answer_class: "no" as const, supports: ["05:10"], conflicts: ["05:00"] },
|
|
{ answer_class: "unsure" as const, supports: [], conflicts: [] },
|
|
],
|
|
information_gain: 0.87,
|
|
source: "dasha_boundary",
|
|
};
|
|
|
|
const EDUCATION_2014_PROBE = {
|
|
id: "p-edu-2014",
|
|
semantic_key: "education.2014",
|
|
candidate_split_hash: "edu-2014-split",
|
|
domain: "education",
|
|
year: 2014,
|
|
question: "2014 年前后有没有升学、转学或换学习环境？",
|
|
candidate_ids: ["05:00", "05:10"],
|
|
expected_outcomes: [
|
|
{ answer_class: "yes" as const, supports: ["05:00"], conflicts: ["05:10"] },
|
|
{ answer_class: "no" as const, supports: ["05:10"], conflicts: ["05:00"] },
|
|
{ answer_class: "unsure" as const, supports: [], conflicts: [] },
|
|
],
|
|
information_gain: 0.64,
|
|
source: "dasha_activation",
|
|
};
|
|
|
|
const FAMILY_2021_COLLECT = {
|
|
year: 2021,
|
|
year_label: "2021 年前后",
|
|
domain: "family",
|
|
event_family: "家人结婚、添丁或住院",
|
|
source: "age_band",
|
|
tracks: ["vimshottari", "narayana"],
|
|
tracks_agree: false,
|
|
unique_minute_claim: false,
|
|
user_meaning: "时间范围锁定 2021 年前后；领域锁定 family；语义目标是家人结婚、添丁或住院。",
|
|
role: "collect",
|
|
phase: "evidence_collection",
|
|
information_gain: 0,
|
|
semantic_key: "family.2021",
|
|
candidate_split_hash: "family:2021",
|
|
candidate_ids: [],
|
|
expected_outcomes: [],
|
|
style_options: STYLE_OPTIONS,
|
|
choice_kind: "existence",
|
|
};
|
|
|
|
const YEARLESS_D24_PROBE = {
|
|
id: "contrast:varga.d24.05:00/05:10",
|
|
semantic_key: "varga.d24.05:00/05:10",
|
|
candidate_split_hash: "varga.d24.05:00/05:10",
|
|
domain: "education",
|
|
year: 0,
|
|
question: "引擎给出的区分机会绑定 D24。",
|
|
candidate_ids: ["05:00", "05:10"],
|
|
expected_outcomes: [
|
|
{ answer_class: "yes" as const, supports: ["05:00"], conflicts: ["05:10"] },
|
|
{ answer_class: "no" as const, supports: ["05:10"], conflicts: ["05:00"] },
|
|
],
|
|
information_gain: 2.5,
|
|
source: "varga_contrast",
|
|
};
|
|
|
|
function eventProbeFromConflict(probe: typeof RELOCATION_2015_PROBE | typeof EDUCATION_2014_PROBE) {
|
|
return {
|
|
year: probe.year,
|
|
year_label: `${probe.year} 年前后`,
|
|
...("month" in probe && probe.month ? { month: probe.month, year_label: `${probe.year} 年 ${probe.month} 月前后` } : {}),
|
|
domain: probe.domain,
|
|
event_family: probe.domain === "education"
|
|
? "升学、转学或换学习环境"
|
|
: "搬家或长期住到外地",
|
|
source: probe.source,
|
|
tracks: ["vimshottari", "narayana"],
|
|
tracks_agree: false,
|
|
unique_minute_claim: false,
|
|
user_meaning: probe.question,
|
|
role: "distinguish",
|
|
phase: "candidate_discriminator",
|
|
information_gain: probe.information_gain,
|
|
semantic_key: probe.semantic_key,
|
|
candidate_split_hash: probe.candidate_split_hash,
|
|
candidate_ids: probe.candidate_ids,
|
|
expected_outcomes: probe.expected_outcomes,
|
|
style_options: STYLE_OPTIONS,
|
|
choice_kind: "existence",
|
|
};
|
|
}
|
|
|
|
function inferenceState() {
|
|
return buildInferenceState({
|
|
range_start: "04:50",
|
|
range_end: "05:10",
|
|
candidates: [
|
|
{ id: "05:00", time: "05:00", relative_support: 10 },
|
|
{ id: "05:10", time: "05:10", relative_support: 10 },
|
|
],
|
|
events: [
|
|
{ id: "e1", domain: "education", year: 2016, precision: "month" },
|
|
{ id: "e2", domain: "career", year: 2018, precision: "year" },
|
|
{ id: "e3", domain: "relationship", year: 2021, precision: "year" },
|
|
{ id: "e4", domain: "family", year: 2023, precision: "year" },
|
|
],
|
|
probes: [{
|
|
id: "p-cd",
|
|
semantic_key: "career.2015",
|
|
candidate_split_hash: "05:00|05:10",
|
|
domain: "career",
|
|
year: 2015,
|
|
question: "2016 年前后有没有高考或重要考试发挥失常？",
|
|
candidate_ids: ["05:00", "05:10"],
|
|
expected_outcomes: [
|
|
{ answer_class: "yes", supports: ["05:00"], conflicts: ["05:10"] },
|
|
{ answer_class: "no", supports: ["05:10"], conflicts: ["05:00"] },
|
|
{ answer_class: "unsure", supports: [], conflicts: [] },
|
|
],
|
|
information_gain: 0.4,
|
|
source: "dasha_boundary",
|
|
}],
|
|
});
|
|
}
|
|
|
|
function choiceDossier() {
|
|
const inference = inferenceState();
|
|
const snapshot = candidateSnapshotFixture();
|
|
Object.assign(snapshot.decision_receipt, { inference_state: inference });
|
|
return dossierFixture({
|
|
latestResult: snapshot,
|
|
conversationSummary: conversationSummaryFixture({
|
|
activeFocus: activeFocusFixture({
|
|
expectedAnswerSchema: {
|
|
choice: {
|
|
prompt: "2016 年前后，有没有明显高考或重要考试发挥失常？",
|
|
option_a: "是，大概就在那段时间",
|
|
option_b: "有类似，但年份不对或不够重大",
|
|
option_c: "没有明显发生",
|
|
option_d: "不记得 / 不确定",
|
|
options: [
|
|
{ key: "A", label: "是，大概就在那段时间", answer_class: "yes" },
|
|
{ key: "B", label: "有类似，但年份不对或不够重大", answer_class: "weak_yes" },
|
|
{ key: "C", label: "没有明显发生", answer_class: "no" },
|
|
{ key: "D", label: "不记得 / 不确定", answer_class: "unsure" },
|
|
],
|
|
},
|
|
probe_id: "p-cd",
|
|
semantic_key: "career.2015",
|
|
},
|
|
}),
|
|
}),
|
|
});
|
|
}
|
|
|
|
function twoProbeInference() {
|
|
return buildInferenceState({
|
|
range_start: "04:45",
|
|
range_end: "05:15",
|
|
candidates: [
|
|
{ id: "05:00", time: "05:00", relative_support: 16 },
|
|
{ id: "05:10", time: "05:10", relative_support: 14 },
|
|
],
|
|
events: [
|
|
{ id: "e1", domain: "career", year: 2020, precision: "month" },
|
|
{ id: "e2", domain: "career", year: 2024, precision: "day" },
|
|
{ id: "e3", domain: "relationship", year: 2024, precision: "month" },
|
|
{ id: "e4", domain: "career", year: 2026, precision: "day" },
|
|
{ id: "e5", domain: "relationship", year: 2024, precision: "day" },
|
|
],
|
|
probes: [RELOCATION_2015_PROBE, EDUCATION_2014_PROBE],
|
|
});
|
|
}
|
|
|
|
function fourEventRows() {
|
|
return [
|
|
{
|
|
id: EVIDENCE_ID,
|
|
source_turn_id: TURN_ID,
|
|
subject: "self",
|
|
event_kind: "career_entry",
|
|
domain: "career",
|
|
occurred_from: "2020-04-01",
|
|
occurred_to: "2020-04-01",
|
|
date_precision: "month",
|
|
summary: "2020 年 4 月开始实习",
|
|
status: "confirmed",
|
|
supersedes_evidence_id: null,
|
|
created_at: "2026-08-28T07:35:45.000Z",
|
|
},
|
|
{
|
|
id: EDUCATION_ID,
|
|
source_turn_id: TURN_ID,
|
|
subject: "self",
|
|
event_kind: "career_entry",
|
|
domain: "career",
|
|
occurred_from: "2024-04-07",
|
|
occurred_to: null,
|
|
date_precision: "day",
|
|
summary: "2024 年 4 月 7 日正式入职",
|
|
status: "confirmed",
|
|
supersedes_evidence_id: null,
|
|
created_at: "2026-08-28T07:36:35.000Z",
|
|
},
|
|
{
|
|
id: RELATIONSHIP_ID,
|
|
source_turn_id: TURN_ID,
|
|
subject: "self",
|
|
event_kind: "relationship_start",
|
|
domain: "relationship",
|
|
occurred_from: "2024-05-01",
|
|
occurred_to: "2024-05-01",
|
|
date_precision: "month",
|
|
summary: "2024 年 5 月认识一位女生",
|
|
status: "confirmed",
|
|
supersedes_evidence_id: null,
|
|
created_at: "2026-08-28T07:37:02.000Z",
|
|
},
|
|
{
|
|
id: CAREER_EXIT_ID,
|
|
source_turn_id: TURN_ID,
|
|
subject: "self",
|
|
event_kind: "career_exit",
|
|
domain: "career",
|
|
occurred_from: "2026-08-11",
|
|
occurred_to: null,
|
|
date_precision: "day",
|
|
summary: "2026 年 8 月 11 日离职",
|
|
status: "confirmed",
|
|
supersedes_evidence_id: null,
|
|
created_at: "2026-08-28T07:36:35.000Z",
|
|
},
|
|
{
|
|
id: RELATIONSHIP_END_ID,
|
|
source_turn_id: TURN_ID,
|
|
subject: "self",
|
|
event_kind: "relationship_end",
|
|
domain: "relationship",
|
|
occurred_from: "2024-08-08",
|
|
occurred_to: "2024-08-08",
|
|
date_precision: "day",
|
|
summary: "2024 年 8 月 8 日与对方分手",
|
|
status: "confirmed",
|
|
supersedes_evidence_id: null,
|
|
created_at: "2026-08-28T07:37:02.000Z",
|
|
},
|
|
];
|
|
}
|
|
|
|
function twoProbeDossier() {
|
|
const inference = twoProbeInference();
|
|
const snapshot = candidateSnapshotFixture({
|
|
decisionReceipt: {
|
|
inference_state: inference,
|
|
discriminating_event_probes: [
|
|
eventProbeFromConflict(RELOCATION_2015_PROBE),
|
|
eventProbeFromConflict(EDUCATION_2014_PROBE),
|
|
],
|
|
},
|
|
});
|
|
return dossierFixture({
|
|
evidenceCount: 5,
|
|
evidence: fourEventRows(),
|
|
latestResult: snapshot,
|
|
conversationSummary: conversationSummaryFixture({
|
|
activeFocus: activeFocusFixture({
|
|
intent: "distinguish_candidates",
|
|
targetDomain: "relocation",
|
|
targetKind: "home_change",
|
|
expectedAnswerSchema: {
|
|
choice: {
|
|
prompt: "2015 年 5 月前后，有没有搬家或长期住到外地？",
|
|
option_a: "明确发生且时间吻合",
|
|
option_b: "发生过但程度较弱",
|
|
option_c: "明确没有发生",
|
|
option_d: "这段记不清楚",
|
|
options: STYLE_OPTIONS.map((option, index) => ({
|
|
key: (["A", "B", "C", "D"] as const)[index]!,
|
|
label: option.label,
|
|
answer_class: option.answer_class,
|
|
})),
|
|
},
|
|
probe_id: RELOCATION_2015_PROBE.id,
|
|
semantic_key: RELOCATION_2015_PROBE.semantic_key,
|
|
candidate_split_hash: RELOCATION_2015_PROBE.candidate_split_hash,
|
|
},
|
|
}),
|
|
}),
|
|
turns: [{
|
|
id: TURN_ID,
|
|
role: "assistant",
|
|
text: "接下来有一个问题需要您点选一下，麻烦看一下下方选项。",
|
|
status: "completed",
|
|
created_at: "2026-08-28T07:36:54.000Z",
|
|
completed_at: "2026-08-28T07:37:34.000Z",
|
|
}],
|
|
});
|
|
}
|
|
|
|
function familyCollectInference() {
|
|
return buildInferenceState({
|
|
range_start: "04:45",
|
|
range_end: "05:15",
|
|
candidates: [
|
|
{ id: "05:00", time: "05:00", relative_support: 16 },
|
|
{ id: "05:10", time: "05:10", relative_support: 14 },
|
|
],
|
|
events: [
|
|
{ id: "e1", domain: "career", year: 2020, precision: "month" },
|
|
{ id: "e2", domain: "career", year: 2024, precision: "day" },
|
|
{ id: "e3", domain: "relationship", year: 2024, precision: "month" },
|
|
{ id: "e4", domain: "career", year: 2026, precision: "day" },
|
|
{ id: "e5", domain: "relationship", year: 2024, precision: "day" },
|
|
],
|
|
probes: [RELOCATION_2015_PROBE, YEARLESS_D24_PROBE],
|
|
});
|
|
}
|
|
|
|
function familyCollectDossier() {
|
|
const inference = familyCollectInference();
|
|
const snapshot = candidateSnapshotFixture({
|
|
decisionReceipt: {
|
|
inference_state: inference,
|
|
discriminating_event_probes: [eventProbeFromConflict(RELOCATION_2015_PROBE)],
|
|
evidence_collection_probes: [FAMILY_2021_COLLECT],
|
|
},
|
|
});
|
|
return dossierFixture({
|
|
evidenceCount: 5,
|
|
evidence: fourEventRows(),
|
|
latestResult: snapshot,
|
|
conversationSummary: conversationSummaryFixture({
|
|
activeFocus: activeFocusFixture({
|
|
intent: "distinguish_candidates",
|
|
targetDomain: "relocation",
|
|
targetKind: "home_change",
|
|
expectedAnswerSchema: {
|
|
choice: {
|
|
prompt: "2015 年 5 月前后，有没有搬家或长期住到外地？",
|
|
option_a: "明确发生且时间吻合",
|
|
option_b: "发生过但程度较弱",
|
|
option_c: "明确没有发生",
|
|
option_d: "这段记不清楚",
|
|
options: STYLE_OPTIONS.map((option, index) => ({
|
|
key: (["A", "B", "C", "D"] as const)[index]!,
|
|
label: option.label,
|
|
answer_class: option.answer_class,
|
|
})),
|
|
},
|
|
probe_id: RELOCATION_2015_PROBE.id,
|
|
semantic_key: RELOCATION_2015_PROBE.semantic_key,
|
|
},
|
|
}),
|
|
}),
|
|
});
|
|
}
|
|
|
|
function persistChoiceAccounting(
|
|
dossier: ReturnType<typeof twoProbeDossier>,
|
|
extra: Parameters<typeof fakeAccounting>[0] = {},
|
|
) {
|
|
return choiceAccounting({
|
|
get_agentic_rectification_case_dossier: () => dossier,
|
|
get_agentic_rectification_case_compute: () => computeFixture(),
|
|
...extra,
|
|
});
|
|
}
|
|
|
|
function choiceAccounting(overrides: Parameters<typeof fakeAccounting>[0] = {}) {
|
|
const receipts = new Map<string, Record<string, unknown>>();
|
|
return fakeAccounting({
|
|
...receiptHandlers,
|
|
get_agentic_rectification_case_dossier: () => choiceDossier(),
|
|
apply_agentic_rectification_choice_action: (_fn, args) => {
|
|
const existing = receipts.get(String(args.p_action_id));
|
|
if (existing) return { ...existing, idempotent: true };
|
|
const receipt = {
|
|
action_id: args.p_action_id,
|
|
status: "applied",
|
|
idempotent: false,
|
|
question_id: args.p_question_id,
|
|
option_id: args.p_option_id,
|
|
probe_id: args.p_inference && typeof args.p_inference === "object"
|
|
? (args.p_inference as { probe_id?: string }).probe_id ?? "p-cd"
|
|
: "p-cd",
|
|
revision: Number(args.p_expected_revision) + 1,
|
|
source_quote: args.p_source_quote,
|
|
derived_context: args.p_derived_context,
|
|
narration: args.p_narration,
|
|
focus_status: args.p_focus_status,
|
|
};
|
|
receipts.set(String(args.p_action_id), receipt);
|
|
return receipt;
|
|
},
|
|
append_agentic_rectification_turn: () => ({ turn_id: TURN_ID, idempotent: false }),
|
|
...overrides,
|
|
});
|
|
}
|
|
|
|
test("does not force-follow when the reader has scrolled upward", () => {
|
|
assert.equal(isNearBottom(1_000, 100, 400), false);
|
|
assert.equal(shouldFollowLatest(false), false);
|
|
assert.equal(isNearBottom(1_000, 520, 400), true);
|
|
assert.equal(shouldFollowLatest(true), true);
|
|
});
|
|
|
|
test("jump-to-latest stays hidden while a choice card still sits in the overlay band", () => {
|
|
assert.equal(shouldShowJumpToLatest(1_000, 100, 400), true);
|
|
assert.equal(shouldShowJumpToLatest(1_000, 520, 400), false);
|
|
assert.equal(shouldShowJumpToLatest(1_000, 500, 400, true), false);
|
|
assert.equal(shouldShowJumpToLatest(1_000, 200, 400, true), true);
|
|
});
|
|
|
|
test("choice quotes come from the option label, not the assistant question year", () => {
|
|
const question = "2016 年前后，有没有明显高考或重要考试发挥失常？";
|
|
assert.equal(quoteIsFromAssistantQuestion("2016年", question), true);
|
|
assert.equal(quoteIsFromAssistantQuestion("是，大概就在那段时间", question), false);
|
|
const card = {
|
|
question_id: QUESTION_ID,
|
|
method_id: "d5",
|
|
prompt: question,
|
|
why: "",
|
|
varga: null,
|
|
choice_mode: "A/B/C/D" as const,
|
|
options: [
|
|
{ key: "A" as const, label: "是，大概就在那段时间", answer_class: "yes" as const, role: "primary" as const },
|
|
{ key: "B" as const, label: "有类似，但年份不对或不够重大", answer_class: "weak_yes" as const, role: "primary" as const },
|
|
{ key: "C" as const, label: "没有明显发生", answer_class: "no" as const, role: "primary" as const },
|
|
{ key: "D" as const, label: "不记得 / 不确定", answer_class: "unsure" as const, role: "secondary" as const },
|
|
],
|
|
stop_label: "先这样，先看当前范围",
|
|
stop_message: "先这样",
|
|
scoring: true,
|
|
probe_id: "p-cd",
|
|
case_revision: 1,
|
|
focus_id: FOCUS_ID,
|
|
};
|
|
assert.equal(userVisibleChoiceLine(card, "A"), "A. 是，大概就在那段时间");
|
|
});
|
|
|
|
test("synthetic choice taps are not kept as chat user lines", () => {
|
|
assert.equal(isStructuredChoiceUserText("A. 发挥明显失常或压力很大"), true);
|
|
assert.equal(isStructuredChoiceUserText("先这样"), true);
|
|
assert.equal(isStructuredChoiceUserText("先这样，先看当前范围"), true);
|
|
assert.equal(isStructuredChoiceUserText("盘外核对（不计分）：A. 明确发生且时间吻合"), true);
|
|
assert.equal(isStructuredChoiceUserText("2016年我上了大学"), false);
|
|
assert.equal(isStructuredChoiceUserText("没有，那年很顺利"), false);
|
|
});
|
|
|
|
test("clicking A applies the choice without invoking a language model", async () => {
|
|
const accounting = choiceAccounting();
|
|
const applied = await applyRectificationChoice(accounting.client, {
|
|
userId: USER_ID,
|
|
caseId: CASE_ID,
|
|
sessionId: SESSION_ID,
|
|
actionId: ACTION_ID,
|
|
action: CHOICE_ACTION,
|
|
focusId: FOCUS_ID,
|
|
questionId: QUESTION_ID,
|
|
probeId: "p-cd",
|
|
optionId: "A",
|
|
expectedRevision: inferenceState().revision,
|
|
});
|
|
assert.equal(applied.applied, true);
|
|
assert.equal(applied.focusId, FOCUS_ID);
|
|
assert.equal(applied.optionId, "A");
|
|
assert.equal(applied.sourceQuote, "是，大概就在那段时间");
|
|
assert.equal(applied.derivedContext.sourceType, "structured_probe_answer");
|
|
assert.deepEqual(applied.derivedContext.referencedDateRange, {
|
|
start: "2015-01-01",
|
|
end: "2015-12-31",
|
|
});
|
|
assert.equal(applied.narration, composeChoiceNarration({
|
|
optionId: "A",
|
|
scoring: true,
|
|
appliedInference: true,
|
|
}));
|
|
const fns = accounting.calls.map((call) => call.fn);
|
|
assert.ok(fns.includes("apply_agentic_rectification_choice_action"));
|
|
assert.equal(fns.includes("append_agentic_rectification_turn"), false);
|
|
assert.equal(fns.includes("record_agentic_rectification_evidence_batch"), false);
|
|
assert.equal(fns.some((fn) => fn === "create_agentic_rectification_run_attempt" || fn.includes("stream")), false);
|
|
const persist = accounting.calls.find((call) => call.fn === "apply_agentic_rectification_choice_action");
|
|
assert.equal(persist?.args.p_focus_id, FOCUS_ID);
|
|
assert.equal(persist?.args.p_focus_status, "resolved");
|
|
assert.equal(persist?.args.p_source_quote, "是，大概就在那段时间");
|
|
assert.notEqual(persist?.args.p_source_quote, "2016年");
|
|
assert.ok(persist?.args.p_inference);
|
|
assert.equal(applied.nextAction.type === "ask_candidate_discriminator"
|
|
|| applied.nextAction.type === "ask_holdout_validation"
|
|
|| applied.nextAction.type === "offer_provisional_range"
|
|
|| applied.nextAction.type === "complete_with_range"
|
|
|| applied.nextAction.type === "ready_to_adopt"
|
|
|| applied.nextAction.type === "ask_fact_collection", true);
|
|
});
|
|
|
|
test("answering a discriminator persists the next dated card so GET still has a tap target", async () => {
|
|
const closed = parseV9CaseDossier(twoProbeDossier());
|
|
assert.ok(closed);
|
|
Object.assign(closed.conversationSummary, { activeFocus: null });
|
|
assert.equal(choiceCardFromCaseDossier(closed), null);
|
|
|
|
const accounting = persistChoiceAccounting(twoProbeDossier(), {
|
|
set_agentic_rectification_conversation_focus: (_fn, args) => ({
|
|
focus: {
|
|
id: NEXT_FOCUS_ID,
|
|
case_id: CASE_ID,
|
|
question_id: args.p_question_id,
|
|
intent: args.p_intent,
|
|
target_evidence_id: args.p_target_evidence_id,
|
|
target_domain: args.p_target_domain,
|
|
target_kind: args.p_target_kind,
|
|
expected_answer_schema: args.p_expected_answer_schema,
|
|
status: "active",
|
|
asked_at: "2026-08-28T07:37:50.000Z",
|
|
resolved_at: null,
|
|
},
|
|
idempotent: false,
|
|
}),
|
|
});
|
|
const applied = await applyRectificationChoice(accounting.client, {
|
|
userId: USER_ID,
|
|
caseId: CASE_ID,
|
|
sessionId: SESSION_ID,
|
|
actionId: ACTION_ID,
|
|
action: CHOICE_ACTION,
|
|
focusId: FOCUS_ID,
|
|
questionId: QUESTION_ID,
|
|
probeId: RELOCATION_2015_PROBE.id,
|
|
optionId: "C",
|
|
expectedRevision: twoProbeInference().revision,
|
|
});
|
|
const setFocus = accounting.calls.find((call) => call.fn === "set_agentic_rectification_conversation_focus");
|
|
assert.ok(setFocus, JSON.stringify(accounting.calls.map((call) => call.fn)));
|
|
assert.equal(setFocus.args.p_intent, "distinguish_candidates");
|
|
const schema = setFocus.args.p_expected_answer_schema as { semantic_key?: string; choice?: { prompt?: string } };
|
|
assert.equal(schema.semantic_key, EDUCATION_2014_PROBE.semantic_key);
|
|
assert.match(schema.choice?.prompt ?? "", /2014/);
|
|
assert.doesNotMatch(schema.choice?.prompt ?? "", /2015/);
|
|
assert.equal(applied.nextInterviewPersisted, true);
|
|
assert.equal(applied.nextChoiceReady, true);
|
|
assert.equal(shouldContinueAfterStructuredChoice(applied.nextAction, applied), false);
|
|
assert.equal(applied.narration, "接下来请点选下面这一问。");
|
|
const turn = accounting.calls.find((call) => call.fn === "append_agentic_rectification_turn");
|
|
assert.equal(turn?.args.p_assistant_message, "接下来请点选下面这一问。");
|
|
|
|
const refreshed = parseV9CaseDossier(twoProbeDossier());
|
|
assert.ok(refreshed);
|
|
const answered = twoProbeInference();
|
|
Object.assign(refreshed, { latestResult: {
|
|
...refreshed.latestResult!,
|
|
decisionReceipt: {
|
|
...(refreshed.latestResult?.decisionReceipt ?? {}),
|
|
inference_state: {
|
|
...answered,
|
|
answered_probes: [{
|
|
probe_id: RELOCATION_2015_PROBE.id,
|
|
semantic_key: RELOCATION_2015_PROBE.semantic_key,
|
|
candidate_split_hash: RELOCATION_2015_PROBE.candidate_split_hash,
|
|
answer_class: "no",
|
|
classified_from: "choice",
|
|
}],
|
|
},
|
|
},
|
|
} });
|
|
Object.assign(refreshed.conversationSummary, { activeFocus: {
|
|
id: NEXT_FOCUS_ID,
|
|
caseId: CASE_ID,
|
|
questionId: String(setFocus.args.p_question_id),
|
|
intent: String(setFocus.args.p_intent),
|
|
targetEvidenceId: null,
|
|
targetDomain: typeof setFocus.args.p_target_domain === "string" ? setFocus.args.p_target_domain : "education",
|
|
targetKind: null,
|
|
expectedAnswerSchema: schema,
|
|
status: "active",
|
|
askedAt: "2026-08-28T07:37:50.000Z",
|
|
resolvedAt: null,
|
|
} });
|
|
const card = choiceCardFromCaseDossier(refreshed);
|
|
assert.ok(card);
|
|
assert.equal(card.focus_id, NEXT_FOCUS_ID);
|
|
assert.match(card.prompt, /2014/);
|
|
assert.doesNotMatch(card.prompt, /2015/);
|
|
});
|
|
|
|
test("answering the last discriminator persists a year-locked family collect, not a yearless D24 card", async () => {
|
|
const accounting = persistChoiceAccounting(familyCollectDossier(), {
|
|
set_agentic_rectification_conversation_focus: () => {
|
|
throw new Error("yearless D24 must not persist a scoring focus");
|
|
},
|
|
});
|
|
const applied = await applyRectificationChoice(accounting.client, {
|
|
userId: USER_ID,
|
|
caseId: CASE_ID,
|
|
sessionId: SESSION_ID,
|
|
actionId: ACTION_ID,
|
|
action: CHOICE_ACTION,
|
|
focusId: FOCUS_ID,
|
|
questionId: QUESTION_ID,
|
|
probeId: RELOCATION_2015_PROBE.id,
|
|
optionId: "C",
|
|
expectedRevision: familyCollectInference().revision,
|
|
});
|
|
assert.equal(
|
|
accounting.calls.some((call) => call.fn === "set_agentic_rectification_conversation_focus"),
|
|
false,
|
|
JSON.stringify(accounting.calls.map((call) => call.fn)),
|
|
);
|
|
assert.equal(applied.nextInterviewPersisted, true);
|
|
assert.equal(applied.nextChoiceReady, false);
|
|
assert.equal(shouldContinueAfterStructuredChoice(applied.nextAction, applied), false);
|
|
assert.match(applied.narration, /2021/);
|
|
assert.match(applied.narration, /家人|结婚|添丁|住院/);
|
|
assert.doesNotMatch(applied.narration, /点选/);
|
|
assert.doesNotMatch(applied.narration, /D24/);
|
|
const turn = accounting.calls.find((call) => call.fn === "append_agentic_rectification_turn");
|
|
assert.match(String(turn?.args.p_assistant_message ?? ""), /2021/);
|
|
});
|
|
|
|
test("replaying the same actionId does not duplicate the applied receipt", async () => {
|
|
const accounting = choiceAccounting();
|
|
const command = {
|
|
userId: USER_ID,
|
|
caseId: CASE_ID,
|
|
sessionId: SESSION_ID,
|
|
actionId: ACTION_ID,
|
|
action: CHOICE_ACTION,
|
|
focusId: FOCUS_ID,
|
|
questionId: QUESTION_ID,
|
|
probeId: "p-cd",
|
|
optionId: "A" as const,
|
|
expectedRevision: inferenceState().revision,
|
|
};
|
|
const first = await applyRectificationChoice(accounting.client, command);
|
|
const second = await applyRectificationChoice(accounting.client, command);
|
|
assert.equal(first.idempotent, false);
|
|
assert.equal(second.idempotent, true);
|
|
assert.equal(
|
|
accounting.calls.filter((call) => call.fn === "apply_agentic_rectification_choice_action").length,
|
|
2,
|
|
);
|
|
const secondCall = accounting.calls.filter((call) => call.fn === "apply_agentic_rectification_choice_action").at(-1);
|
|
assert.equal(secondCall?.args.p_action_id, ACTION_ID);
|
|
});
|
|
|
|
test("keeps the applied answer when narration persistence fails", async () => {
|
|
const accounting = choiceAccounting({
|
|
append_agentic_rectification_turn: () => {
|
|
throw new Error("turn_write_failed");
|
|
},
|
|
});
|
|
const applied = await applyRectificationChoice(accounting.client, {
|
|
userId: USER_ID,
|
|
caseId: CASE_ID,
|
|
sessionId: SESSION_ID,
|
|
actionId: ACTION_ID,
|
|
action: CHOICE_ACTION,
|
|
focusId: FOCUS_ID,
|
|
questionId: QUESTION_ID,
|
|
probeId: "p-cd",
|
|
optionId: "A",
|
|
expectedRevision: inferenceState().revision,
|
|
});
|
|
assert.equal(applied.applied, true);
|
|
assert.equal(applied.status, "applied");
|
|
assert.equal(applied.narrationPersisted, false);
|
|
assert.match(applied.narration, /已记录你的选择/);
|
|
});
|
|
|
|
test("stop_and_review does not write an inference transition", async () => {
|
|
const accounting = choiceAccounting();
|
|
const applied = await applyRectificationChoice(accounting.client, {
|
|
userId: USER_ID,
|
|
caseId: CASE_ID,
|
|
sessionId: SESSION_ID,
|
|
actionId: ACTION_ID,
|
|
action: STOP_ACTION,
|
|
focusId: FOCUS_ID,
|
|
questionId: QUESTION_ID,
|
|
optionId: "stop",
|
|
expectedRevision: inferenceState().revision,
|
|
});
|
|
assert.equal(applied.optionId, "stop");
|
|
const persist = accounting.calls.find((call) => call.fn === "apply_agentic_rectification_choice_action");
|
|
assert.equal(persist?.args.p_inference, null);
|
|
assert.equal(persist?.args.p_focus_status, "skipped");
|
|
assert.ok(accounting.calls.some((call) => call.fn === "append_agentic_rectification_turn"));
|
|
});
|
|
|
|
test("turn_decision stays inside the configured byte budget", () => {
|
|
const dossier = parseV9CaseDossier(choiceDossier());
|
|
assert.ok(dossier);
|
|
const projection = projectTurnDecision(dossier);
|
|
assert.equal(projection.projection, "turn_decision");
|
|
const currentQuestion = projection.current_question as { prompt?: string } | null;
|
|
assert.equal(currentQuestion?.prompt, "2016 年前后，有没有明显高考或重要考试发挥失常？");
|
|
assert.ok(turnDecisionByteLength(projection) <= TURN_DECISION_MAX_BYTES);
|
|
assert.ok(!("birth_context" in projection));
|
|
assert.ok(!("baseline_birth_snapshot" in projection));
|
|
});
|
|
|
|
test("unrenderable focus schema stays visible as current_question, not null", () => {
|
|
const snapshot = candidateSnapshotFixture();
|
|
Object.assign(snapshot.decision_receipt, { inference_state: inferenceState() });
|
|
const dossier = parseV9CaseDossier(dossierFixture({
|
|
latestResult: snapshot,
|
|
conversationSummary: conversationSummaryFixture({
|
|
activeFocus: activeFocusFixture({
|
|
expectedAnswerSchema: { choice: { prompt: "坏题" } },
|
|
}),
|
|
}),
|
|
}));
|
|
assert.ok(dossier);
|
|
const projection = projectTurnDecision(dossier);
|
|
const currentQuestion = projection.current_question as {
|
|
unrenderable?: boolean;
|
|
reason?: string;
|
|
prompt?: string | null;
|
|
} | null;
|
|
assert.equal(currentQuestion?.unrenderable, true);
|
|
assert.equal(currentQuestion?.reason, "invalid_choice_schema");
|
|
assert.equal(currentQuestion?.prompt, null);
|
|
assert.equal(projection.current_probe, null);
|
|
});
|
|
|
|
test("collection focus without choice copy is not an unrenderable current_question", () => {
|
|
const snapshot = candidateSnapshotFixture();
|
|
Object.assign(snapshot.decision_receipt, { inference_state: inferenceState() });
|
|
const dossier = parseV9CaseDossier(dossierFixture({
|
|
latestResult: snapshot,
|
|
conversationSummary: conversationSummaryFixture({
|
|
activeFocus: activeFocusFixture({
|
|
intent: "clarify_event_date",
|
|
expectedAnswerSchema: { required: ["month"] },
|
|
}),
|
|
}),
|
|
}));
|
|
assert.ok(dossier);
|
|
const projection = projectTurnDecision(dossier);
|
|
assert.equal(projection.current_question, null);
|
|
});
|
|
|
|
test("turn_decision hides current_probe unless a valid current_question exists", () => {
|
|
const withFocus = projectTurnDecision(parseV9CaseDossier(choiceDossier())!);
|
|
assert.ok(withFocus.current_question);
|
|
assert.ok(withFocus.current_probe);
|
|
assert.equal((withFocus.question_contract as { version?: string }).version, "probe-question-v1");
|
|
|
|
const snapshot = candidateSnapshotFixture();
|
|
Object.assign(snapshot.decision_receipt, { inference_state: inferenceState() });
|
|
const withoutFocus = projectTurnDecision(parseV9CaseDossier(dossierFixture({
|
|
latestResult: snapshot,
|
|
conversationSummary: conversationSummaryFixture({ activeFocus: null }),
|
|
}))!);
|
|
assert.equal(withoutFocus.current_question, null);
|
|
assert.equal(withoutFocus.current_probe, null);
|
|
assert.equal((withoutFocus.inference as { next_probe?: unknown } | null)?.next_probe, null);
|
|
});
|
|
|
|
test("truncated and timed-out runs return concrete finish reasons", () => {
|
|
assert.equal(mapModelFinishToErrorCode({
|
|
finishReason: "length",
|
|
aborted: false,
|
|
timedOut: false,
|
|
answerText: "部分回答",
|
|
stepCount: 1,
|
|
maxSteps: 8,
|
|
}), "answer_truncated");
|
|
assert.equal(mapModelFinishToErrorCode({
|
|
finishReason: "stop",
|
|
aborted: true,
|
|
timedOut: true,
|
|
answerText: "",
|
|
stepCount: 1,
|
|
maxSteps: 8,
|
|
}), "run_timeout");
|
|
assert.equal(userFacingRunFailure("answer_truncated"), "模型输出达到上限，状态已记录。");
|
|
assert.equal(userFacingRunFailure("run_timeout"), "服务端运行超时，状态已记录。");
|
|
assert.equal(isIncompleteRunBanner("本轮处理未完成，请稍后再试"), true);
|
|
assert.equal(isIncompleteRunBanner("已记录你的选择，并更新了候选比较。"), false);
|
|
});
|
|
|
|
test("structured choice only continues through the agent when another question is required", () => {
|
|
for (const type of [
|
|
"ask_fact_collection",
|
|
"ask_candidate_discriminator",
|
|
"ask_holdout_validation",
|
|
]) {
|
|
assert.equal(shouldContinueAfterStructuredChoice({ type }), true);
|
|
}
|
|
for (const type of [
|
|
"offer_provisional_range",
|
|
"complete_with_range",
|
|
"ready_to_adopt",
|
|
]) {
|
|
assert.equal(shouldContinueAfterStructuredChoice({ type }), false);
|
|
}
|
|
assert.equal(shouldContinueAfterStructuredChoice(null), false);
|
|
assert.equal(
|
|
shouldContinueAfterStructuredChoice({ type: "ask_candidate_discriminator" }, { nextInterviewPersisted: true }),
|
|
false,
|
|
);
|
|
});
|
|
|
|
test("structured choice narration never persists a fake loading state", () => {
|
|
for (const narration of [
|
|
composeChoiceNarration({ optionId: "A", scoring: true, appliedInference: true }),
|
|
composeChoiceNarration({ optionId: "A", scoring: true, appliedInference: false }),
|
|
composeChoiceNarration({ optionId: "A", scoring: false, appliedInference: false }),
|
|
]) {
|
|
assert.doesNotMatch(narration, /正在准备下一步/);
|
|
}
|
|
});
|
|
|
|
test("the public agent route treats structured choice as a non-model command", () => {
|
|
const route = readFileSync(new URL("../src/app/api/rectification/agent/route.ts", import.meta.url), "utf8");
|
|
const start = route.indexOf("if (isStructuredChoice)");
|
|
const end = route.indexOf("const selectedModel", start);
|
|
assert.ok(start >= 0 && end > start);
|
|
const block = route.slice(start, end);
|
|
assert.match(block, /applyRectificationChoice\(accounting/);
|
|
assert.doesNotMatch(block, /runV9AgentTurn/);
|
|
assert.doesNotMatch(block, /getRectificationV9Agent/);
|
|
assert.doesNotMatch(block, /authorizeUsage/);
|
|
assert.match(route, /"answer_choice"/);
|
|
assert.match(route, /"stop_and_review"/);
|
|
assert.match(route, /export const maxDuration = 240/);
|
|
const chat = readFileSync(new URL("../src/components/rectification-agentic-chat.tsx", import.meta.url), "utf8");
|
|
assert.match(chat, /isPersistedFocusId\(focusId\)/);
|
|
assert.match(chat, /focusId,/);
|
|
assert.match(chat, /shouldContinueAfterStructuredChoice\(payload\?\.nextAction, payload\)/);
|
|
assert.match(chat, /send\("read_only", ""\)/);
|
|
assert.match(chat, /willContinue/);
|
|
assert.match(chat, /current\.filter\(\(message\) => message\.renderKey !== assistantRenderKey\)/);
|
|
assert.match(chat, /回到最新/);
|
|
assert.match(chat, /followTailRef\.current/);
|
|
});
|
|
|
|
test("rectification attempt timeout stays under the agent route budget", () => {
|
|
const agentRun = readFileSync(new URL("../src/lib/rectification-agentic/v9/agent-run.ts", import.meta.url), "utf8");
|
|
const regenerate = readFileSync(
|
|
new URL("../src/app/api/rectification/cases/[caseId]/turns/[turnId]/regenerate/route.ts", import.meta.url),
|
|
"utf8",
|
|
);
|
|
assert.match(agentRun, /RECTIFICATION_AGENT_ATTEMPT_TIMEOUT_MS = 210_000/);
|
|
assert.match(agentRun, /RECTIFICATION_AGENT_ROUTE_MAX_DURATION_S = 240/);
|
|
assert.match(regenerate, /export const maxDuration = 240/);
|
|
assert.ok(210_000 < 240_000);
|
|
assert.match(agentRun, /RETRYABLE_ERROR_CODES = new Set\(\[/);
|
|
const retryable = agentRun.slice(
|
|
agentRun.indexOf("const RETRYABLE_ERROR_CODES"),
|
|
agentRun.indexOf("function streamFinishReason"),
|
|
);
|
|
assert.match(agentRun, /state_invariant_failed/);
|
|
assert.doesNotMatch(retryable, /state_invariant_failed/);
|
|
assert.doesNotMatch(agentRun, /stale_question/);
|
|
assert.doesNotMatch(agentRun, /revision_conflict/);
|
|
});
|
|
|
|
test("derived questionId is an audit label, not the choice identity", async () => {
|
|
const accounting = choiceAccounting();
|
|
const applied = await applyRectificationChoice(accounting.client, {
|
|
userId: USER_ID,
|
|
caseId: CASE_ID,
|
|
sessionId: SESSION_ID,
|
|
actionId: ACTION_ID,
|
|
action: CHOICE_ACTION,
|
|
focusId: FOCUS_ID,
|
|
questionId: "derived-from-stem:not-the-db-row",
|
|
probeId: "p-cd",
|
|
optionId: "A",
|
|
expectedRevision: inferenceState().revision,
|
|
});
|
|
assert.equal(applied.applied, true);
|
|
assert.equal(applied.focusId, FOCUS_ID);
|
|
const persist = accounting.calls.find((call) => call.fn === "apply_agentic_rectification_choice_action");
|
|
assert.equal(persist?.args.p_focus_id, FOCUS_ID);
|
|
assert.equal(persist?.args.p_question_id, QUESTION_ID);
|
|
});
|
|
|
|
test("a mismatched focusId is stale_question even when questionId matches", async () => {
|
|
await assert.rejects(
|
|
() => applyRectificationChoice(choiceAccounting().client, {
|
|
userId: USER_ID,
|
|
caseId: CASE_ID,
|
|
sessionId: SESSION_ID,
|
|
actionId: ACTION_ID,
|
|
action: CHOICE_ACTION,
|
|
focusId: "cdcdcdcd-cdcd-4dcd-8dcd-cdcdcdcdcdcd",
|
|
questionId: QUESTION_ID,
|
|
probeId: "p-cd",
|
|
optionId: "A",
|
|
expectedRevision: inferenceState().revision,
|
|
}),
|
|
(error: unknown) => {
|
|
assert.ok(error instanceof RectificationToolServiceError);
|
|
assert.match(error.message, /stale_question/);
|
|
return true;
|
|
},
|
|
);
|
|
});
|
|
|
|
test("choice identity SQL keys stale_question to inactive focus, not question_id", () => {
|
|
const migration = readFileSync(
|
|
new URL("../supabase/migrations/20260826020000_rectification_choice_focus_identity.sql", import.meta.url),
|
|
"utf8",
|
|
);
|
|
assert.match(migration, /v_focus\.status is distinct from 'active'/);
|
|
assert.doesNotMatch(migration, /v_focus\.question_id is distinct from/);
|
|
assert.doesNotMatch(migration, /p_question_id is distinct from v_focus\.question_id/);
|
|
const route = readFileSync(new URL("../src/app/api/rectification/agent/route.ts", import.meta.url), "utf8");
|
|
assert.match(route, /focusId: z\.string\(\)\.uuid\(\)/);
|
|
assert.match(route, /选择题请求缺少 actionId、focusId 或 expectedRevision/);
|
|
});
|