@zhushanwen/pi-subagent-workflow 8.6.0 → 8.8.0

package/src/execution/__tests__/format-schema-instruction.test.ts

@@ -1,169 +0,0 @@
-// src/execution/__tests__/format-schema-instruction.test.ts
-//
-// 锁定 formatSchemaInstruction 契约：构造 schema enforcement 的 MANDATORY 指令。
-// [审查项#2] 注入链变更：指令仅经 agent-opts-resolver 进 appendSystemPrompt
-// （ASP 单点），runSpawn 的 task + instruction 后缀拼接已删除——本测试随之改锁
-// resolver 导出的同名函数。一旦漏掉 "structured-output" 关键词或 JSON 序列化漂移，
-// schema 模式会静默失效——agent 可能直接把 JSON 写进文本响应。
-//
-// [审查项#4] 锁定 AP 告知句：注入侧校验用 additionalProperties:false 收窄后的
-// parameters，指令必须前置告知「schema 外字段被拒」，否则拒绝显得凭空。
-//
-// 纯函数测试：不依赖 Pi 运行时、不 spawn 进程、不 mock。只 import 被测函数。
-import { describe, expect, it } from "vitest";
-
-import { formatSchemaInstruction } from "../../orchestration/agent-opts-resolver.ts";
-
-describe("formatSchemaInstruction", () => {
-  // ── 指令文本契约 ──────────────────────────────────────────────
-
-  it("contains the structured-output tool keyword", () => {
-    const out = formatSchemaInstruction({ type: "object" });
-    expect(out).toContain("structured-output");
-  });
-
-  it("emits a MANDATORY structured-output directive (not free-form JSON)", () => {
-    const out = formatSchemaInstruction({ type: "object" });
-    expect(out).toContain("MANDATORY");
-    expect(out).toContain("must be calling the `structured-output` tool");
-    expect(out).toContain("Do NOT output JSON in your text response");
-  });
-
-  it("announces that fields outside the schema are rejected (AP 告知，审查项#4)", () => {
-    const out = formatSchemaInstruction({ type: "object" });
-    expect(out).toContain("Fields not defined in this schema are rejected");
-    expect(out).toContain("do not add extra fields");
-  });
-
-  it("author-declared additionalProperties → weak wording (AP 条件化：声明时作者 schema 管辖额外字段)", () => {
-    // D4 只在「未声明」时注入 false；作者显式声明 true 时额外字段实际放行——
-    // 无条件「一律拒绝」强承诺与参数层行为不符，改用弱承诺措辞。
-    const out = formatSchemaInstruction({ type: "object", additionalProperties: true });
-    expect(out).not.toContain("Fields not defined in this schema are rejected");
-    expect(out).toContain("follow this schema's own additionalProperties declaration");
-  });
-
-  // ── schema 序列化（compact：与 schemaEnv 复用同串，IF7 #13）───────
-
-  it("embeds the schema as compact JSON inside a fenced block", () => {
-    const schema: Record<string, unknown> = {
-      type: "object",
-      properties: { name: { type: "string" } },
-    };
-    const out = formatSchemaInstruction(schema);
-    // 必须包含 JSON.stringify(schema) 的完整结果（compact，无缩进）
-    expect(out).toContain(JSON.stringify(schema));
-    expect(out).toContain("```json");
-    expect(out).toContain("```");
-  });
-
-  it("locks the full output structure for a minimal schema", () => {
-    const out = formatSchemaInstruction({ type: "object" });
-    // 完整结构快照——任何指令措辞/顺序/序列化漂移都会被捕获。
-    // [HISTORICAL] runner task 后缀双重注入删除后，本函数迁至 resolver 成为唯一
-    // 文案源；JSON 从 pretty（indent=2）改为 compact（与 schemaEnv 同串复用）。
-    expect(out).toBe(
-      [
-        "## MANDATORY: Structured Output Requirement",
-        "",
-        "This task requires structured output.",
-        "Your FINAL action must be calling the `structured-output` tool.",
-        "",
-        "Your call arguments ARE the result data itself — the tool's parameter schema IS the required shape of your result.",
-        "Your result must conform to this schema:",
-        "```json",
-        '{"type":"object"}',
-        "```",
-        "",
-        "Rules:",
-        "- Call the structured-output tool with your result data as its arguments. The system validates them against the schema above automatically.",
-        "- Do NOT output JSON in your text response — use the structured-output tool.",
-        "- Do NOT skip this step. The structured-output call IS your result.",
-        "- Complete all other work FIRST, then call structured-output as the last action.",
-        "- Fields not defined in this schema are rejected — do not add extra fields.",
-      ].join("\n"),
-    );
-  });
-
-  it("locks the single-parameter wording: arguments ARE the data coexists with the mandatory tool call", () => {
-    const out = formatSchemaInstruction({ type: "object" });
-    // 新口径双支柱必须共存：① 参数即数据（arguments ARE the data）② 必须调用工具
-    expect(out).toContain("arguments ARE the result data itself");
-    expect(out).toContain("parameter schema IS the required shape of your result");
-    expect(out).toContain("must be calling the `structured-output` tool");
-    // 旧双参数警告语义不得回流（工具已是单参数形态，警告失去对象）
-    expect(out).not.toMatch(/do NOT pass a .schema. parameter/i);
-    expect(out).not.toContain("ONLY the `data` parameter");
-    expect(out).not.toContain("enforced by the system");
-  });
-
-  // ── 特殊字符转义（注入风险路径）──────────────────────────────
-
-  it("escapes double quotes inside schema string values", () => {
-    const schema: Record<string, unknown> = { prompt: 'say "hi"' };
-    const out = formatSchemaInstruction(schema);
-    // JSON.stringify 会把内层 " 转义为 \"
-    expect(out).toContain('say \\"hi\\"');
-    // 原始未转义形式（含成对字面双引号）绝不能回流进 JSON 体内
-    expect(out).not.toContain('say "hi"');
-  });
-
-  it("escapes newlines inside schema string values", () => {
-    const schema: Record<string, unknown> = { text: "line1\nline2" };
-    const out = formatSchemaInstruction(schema);
-    // 换行被序列化为字面反斜杠-n，不能是真实换行符
-    expect(out).toContain("line1\\nline2");
-    expect(out).not.toContain("line1\nline2");
-  });
-
-  it("escapes backslashes inside schema string values", () => {
-    const schema: Record<string, unknown> = { path: "C:\\Users\\x" };
-    const out = formatSchemaInstruction(schema);
-    // 单反斜杠被序列化为 \\，避免后续解析误把转义序列当指令
-    expect(out).toContain("C:\\\\Users\\\\x");
-    expect(out).not.toContain("C:\\Users\\x");
-  });
-
-  // ── 边界值 ────────────────────────────────────────────────────
-
-  it("handles empty schema object", () => {
-    const out = formatSchemaInstruction({});
-    expect(out).toContain("structured-output");
-    expect(out).toContain("{}");
-  });
-
-  it("preserves null values in schema", () => {
-    const schema: Record<string, unknown> = { default: null };
-    const out = formatSchemaInstruction(schema);
-    // JSON.stringify 对 null 保留字面 "null"（不会 omit 键，也不会变字符串）。
-    // compact 序列化无空格：'"default":null'
-    expect(out).toContain('"default":null');
-  });
-
-  it("serializes nested objects and arrays", () => {
-    const schema: Record<string, unknown> = {
-      type: "object",
-      required: ["name", "age"],
-      properties: {
-        name: { type: "string" },
-        age: { type: "integer", minimum: 0 },
-      },
-    };
-    const out = formatSchemaInstruction(schema);
-    expect(out).toContain(JSON.stringify(schema));
-  });
-
-  // ── 确定性 ────────────────────────────────────────────────────
-
-  it("is deterministic — same schema produces identical output", () => {
-    const schema: Record<string, unknown> = { a: 1, b: [2, 3] };
-    expect(formatSchemaInstruction(schema)).toBe(formatSchemaInstruction(schema));
-  });
-
-  it("is deterministic across different object key insertion (value-equal schemas)", () => {
-    // JSON.stringify 按对象自身属性顺序序列化；同序构造的等价 schema 应产出相同指令
-    const a: Record<string, unknown> = { x: 1, y: 2 };
-    const b: Record<string, unknown> = { x: 1, y: 2 };
-    expect(formatSchemaInstruction(a)).toBe(formatSchemaInstruction(b));
-  });
-});

package/src/execution/__tests__/format.test.ts

@@ -1,458 +0,0 @@
-// src/__tests__/format.test.ts
-import { visibleWidth } from "@earendil-works/pi-tui";
-import { describe, expect, it } from "vitest";
-
-import {
-  formatElapsedSeconds,
-  formatTokens,
-  padToVisible,
-  sanitizeLabel,
-  segFillColored,
-  shortId,
-  spinnerGlyph,
-  statusGlyph,
-  truncLine,
-  wrapText,
-} from "../../interface/format.ts";
-
-// ============================================================
-// formatTokens
-// ============================================================
-describe("formatTokens", () => {
-  it("shows plain value below 1000", () => {
-    expect(formatTokens(0)).toBe("0");
-    expect(formatTokens(820)).toBe("820");
-    expect(formatTokens(999)).toBe("999");
-  });
-
-  it("shows N.Nk between 1000 and 9999", () => {
-    expect(formatTokens(1000)).toBe("1.0k");
-    expect(formatTokens(8200)).toBe("8.2k");
-    expect(formatTokens(9999)).toBe("10.0k");
-  });
-
-  it("shows rounded Nk at 10000+", () => {
-    expect(formatTokens(10000)).toBe("10k");
-    expect(formatTokens(23000)).toBe("23k");
-    expect(formatTokens(99999)).toBe("100k");
-  });
-});
-
-// ============================================================
-// formatElapsedSeconds
-// ============================================================
-describe("formatElapsedSeconds", () => {
-  it("shows Xs below 60", () => {
-    expect(formatElapsedSeconds(0)).toBe("0s");
-    expect(formatElapsedSeconds(12)).toBe("12s");
-    expect(formatElapsedSeconds(59)).toBe("59s");
-  });
-
-  it("shows Xm Ys between 60 and 3599", () => {
-    expect(formatElapsedSeconds(60)).toBe("1m0s");
-    expect(formatElapsedSeconds(72)).toBe("1m12s");
-    expect(formatElapsedSeconds(3599)).toBe("59m59s");
-  });
-
-  it("shows Xh Ym at 3600+", () => {
-    expect(formatElapsedSeconds(3600)).toBe("1h0m");
-    expect(formatElapsedSeconds(3661)).toBe("1h1m");
-    expect(formatElapsedSeconds(7325)).toBe("2h2m");
-  });
-});
-
-// ============================================================
-// statusGlyph
-// ============================================================
-describe("statusGlyph", () => {
-  it("running → no icon, accent color", () => {
-    expect(statusGlyph("running")).toEqual({ icon: undefined, color: "accent" });
-  });
-
-  it("done → checkmark, success", () => {
-    expect(statusGlyph("closed")).toEqual({ icon: "✓", color: "success" });
-  });
-
-  it("closed → checkmark, success", () => {
-    expect(statusGlyph("closed")).toEqual({ icon: "✓", color: "success" });
-  });
-});
-
-// ============================================================
-// spinnerGlyph
-// ============================================================
-describe("spinnerGlyph", () => {
-  it("returns a frame for valid seed", () => {
-    expect(spinnerGlyph(0)).toBe("⠋");
-    expect(spinnerGlyph(1)).toBe("⠙");
-    expect(spinnerGlyph(9)).toBe("⠏");
-  });
-
-  it("wraps around (mod 10)", () => {
-    expect(spinnerGlyph(10)).toBe("⠋");
-    expect(spinnerGlyph(15)).toBe("⠴"); // index 5
-  });
-
-  it("falls back to frame 0 on NaN", () => {
-    expect(spinnerGlyph(NaN)).toBe("⠋");
-  });
-
-  it("handles negative seeds via abs", () => {
-    expect(spinnerGlyph(-1)).toBe("⠙"); // abs(-1) % 10 = 1 → ⠙
-    expect(spinnerGlyph(-10)).toBe("⠋");
-  });
-
-  it("falls back to frame 0 on Infinity", () => {
-    expect(spinnerGlyph(Infinity)).toBe("⠋");
-  });
-});
-
-// ============================================================
-// sanitizeLabel
-// ============================================================
-describe("sanitizeLabel", () => {
-  it("replaces CRLF/LF with single space", () => {
-    expect(sanitizeLabel("line1\r\nline2\nline3")).toBe("line1 line2 line3");
-  });
-
-  it("replaces tabs with 2 spaces", () => {
-    expect(sanitizeLabel("a\tb")).toBe("a  b");
-  });
-
-  it("collapses multiple consecutive newlines into one space", () => {
-    // /[\r\n]+/g treats \r\n\r\n as one match → single space
-    expect(sanitizeLabel("a\r\n\r\nb")).toBe("a b");
-  });
-
-  it("leaves clean text unchanged", () => {
-    expect(sanitizeLabel("read foo.ts")).toBe("read foo.ts");
-  });
-});
-
-// ============================================================
-// padToVisible
-// ============================================================
-describe("padToVisible", () => {
-  it("pads short text to width with trailing spaces", () => {
-    expect(padToVisible("ab", 5)).toBe("ab   ");
-  });
-
-  it("returns unchanged when already at width", () => {
-    expect(padToVisible("hello", 5)).toBe("hello");
-  });
-
-  it("returns unchanged when wider than width", () => {
-    expect(padToVisible("hello world", 5)).toBe("hello world");
-  });
-
-  it("handles CJK width (2 columns per char)", () => {
-    // 你好 = 4 visible columns
-    expect(padToVisible("你好", 6)).toBe("你好  ");
-  });
-});
-
-// ============================================================
-// segFillColored
-// ============================================================
-describe("segFillColored", () => {
-  it("returns empty string for width <= 0", () => {
-    expect(segFillColored("title", "-", 0)).toBe("");
-    expect(segFillColored("title", "-", -1)).toBe("");
-  });
-
-  it("pure fill when no title", () => {
-    expect(segFillColored(undefined, "-", 5)).toBe("-----");
-  });
-
-  it("title + fill to width", () => {
-    expect(segFillColored("Hi", "-", 5)).toBe("Hi---");
-  });
-
-  it("truncates title when wider than width", () => {
-    const result = segFillColored("Hello World", "-", 5);
-    // title visible width 11 > 5 → truncated to 5 (with ellipsis = 4 chars + …)
-    expect(result.length).toBeLessThanOrEqual(10); // visible width 5 but may include ANSI
-  });
-
-  it("preserves ANSI in title and fill separately (no nesting color loss)", () => {
-    const redTitle = "\x1b[31mHi\x1b[0m";
-    const blueFill = "\x1b[34m-\x1b[0m";
-    const result = segFillColored(redTitle, blueFill, 5);
-    // title visible width = 2 ("Hi"), fill count = 3
-    expect(result).toContain(redTitle);
-    expect(result).toContain(blueFill);
-    // fill repeated 3 times
-    expect(result).toBe(redTitle + blueFill + blueFill + blueFill);
-  });
-});
-
-// ============================================================
-// truncLine
-// ============================================================
-describe("truncLine", () => {
-  it("returns text unchanged when within width", () => {
-    expect(truncLine("hello", 10)).toBe("hello");
-    expect(truncLine("hello", 5)).toBe("hello");
-  });
-
-  it("returns empty string for width <= 0", () => {
-    expect(truncLine("hello", 0)).toBe("");
-  });
-
-  it("truncates with ellipsis when exceeding width", () => {
-    const result = truncLine("hello world", 8);
-    // 纯文本截断不发 \x1b[0m（全局重置会破坏外层背景色）
-    expect(result.endsWith("…")).toBe(true);
-    expect(result).not.toContain("\x1b[0m");
-    // visible width should be 8 (7 chars + ellipsis)
-  });
-
-  it("handles CJK characters (2 columns each)", () => {
-    // 你好世界 = 8 visible columns; truncate to 5 → 2 chars (4 cols) + …
-    const result = truncLine("你好世界", 5);
-    expect(result.endsWith("…")).toBe(true);
-    expect(result).not.toContain("\x1b[0m");
-  });
-
-  it("handles emoji correctly", () => {
-    const result = truncLine("😀😁😂🤣😃", 3);
-    expect(result.endsWith("…")).toBe(true);
-    expect(result).not.toContain("\x1b[0m");
-  });
-
-  it("reapplies active ANSI styles before ellipsis (no background break)", () => {
-    // red text that exceeds width → ellipsis should have red re-applied
-    const input = "\x1b[31mhello world this is long\x1b[0m";
-    const result = truncLine(input, 10);
-    expect(result.endsWith("…\x1b[0m")).toBe(true);
-    // The ellipsis should be preceded by the active red style (re-applied)
-    // Check that the last grapheme sequence includes the red SGR before …
-    expect(result).toMatch(/\x1b\[31m…\x1b\[0m$/);
-  });
-
-  it("clears style stack on reset code", () => {
-    // text with reset in the middle → after reset, no style re-applied
-    const input = "\x1b[31mab\x1b[0mcdefghijk";
-    const result = truncLine(input, 6);
-    // reset 后 activeStyles 为空 → 截断不发 \x1b[0m
-    expect(result.endsWith("…")).toBe(true);
-  });
-
-  it("flattens newlines to spaces (single-line rendering safety)", () => {
-    // 多行 prompt / turn.text 含 \n，单行渲染时 \n 会意外换行破坏行对齐。
-    // truncLine 作为单行渲染入口必须剥离 \n（用空格替代保留词边界）。
-    const multiLine = "只读任务：分析核心逻辑。\n项目根：/Users/test\n输出：报告";
-    const result = truncLine(multiLine, 80);
-    expect(result).not.toContain("\n");
-    expect(result).toContain("只读任务");
-    expect(result).toContain("项目根");
-  });
-
-  it("flattens \r\n sequences", () => {
-    const crlf = "line1\r\nline2\r\nline3";
-    const result = truncLine(crlf, 80);
-    expect(result).not.toContain("\r");
-    expect(result).not.toContain("\n");
-  });
-});
-
-// ============================================================
-// truncLine — legacy 对拍（indexOf 化重构的逐字节等价锚定）
-// ============================================================
-
-/**
- * 旧实现参照：改造前的 per-char 版本，逐字取自 git 历史（093e28fe3~1 的
- * format.ts truncLine）。现行实现（indexOf("\\x1b") 段跳过）必须与其逐字节
- * 等价——本函数即对拍 oracle，与 helpers-bounded-serialize.test.ts 的
- * legacySerialize 同形态（参照物随测试落盘，防「验证只存在于改造当时」）。
- */
-const legacySegmenter = new Intl.Segmenter(undefined, { granularity: "grapheme" });
-
-function legacyTruncLine(text: string, maxWidth: number): string {
-  if (maxWidth <= 0) return "";
-  const flat = text.replace(/[\r\n]+/g, " ");
-  if (visibleWidth(flat) <= maxWidth) return flat;
-
-  const targetWidth = Math.max(0, maxWidth - 1);
-  let result = "";
-  let currentWidth = 0;
-  let activeStyles: string[] = [];
-  let i = 0;
-
-  while (i < flat.length) {
-    // 捕获 ANSI SGR 序列
-    const ansiMatch = flat.slice(i).match(/^\x1b\[[0-9;]*m/);
-    if (ansiMatch) {
-      const code = ansiMatch[0];
-      result += code;
-
-      if (code === "\x1b[0m" || code === "\x1b[m") {
-        activeStyles = []; // reset → 清空栈
-      } else {
-        activeStyles.push(code);
-      }
-      i += code.length;
-      continue;
-    }
-
-    // 找到下一段纯文本(非 ANSI)的边界
-    let end = i;
-    while (end < flat.length && !flat.slice(end).match(/^\x1b\[[0-9;]*m/)) {
-      end++;
-    }
-
-    // 按 grapheme 迭代这段文本,累加到 targetWidth
-    const textPortion = flat.slice(i, end);
-    for (const seg of legacySegmenter.segment(textPortion)) {
-      const grapheme = seg.segment;
-      const graphemeWidth = visibleWidth(grapheme);
-
-      if (currentWidth + graphemeWidth > targetWidth) {
-        return result + activeStyles.join("") + "…" + (activeStyles.length ? "\x1b[0m" : "");
-      }
-
-      result += grapheme;
-      currentWidth += graphemeWidth;
-    }
-    i = end;
-  }
-
-  // 理论上 visibleWidth 检查已提前返回,此行兜底
-  return result + activeStyles.join("") + "…" + (activeStyles.length ? "\x1b[0m" : "");
-}
-
-/** 确定性 PRNG（mulberry32）：fuzz 用例固定 seed 生成，失败可按 case 编号 + seed 复现。 */
-function mulberry32(seed: number): () => number {
-  let s = seed | 0;
-  return () => {
-    s = (s + 0x6d2b79f5) | 0;
-    let t = Math.imul(s ^ (s >>> 15), 1 | s);
-    t = (t + Math.imul(t ^ (t >>> 7), 61 | t)) ^ t;
-    return ((t ^ (t >>> 14)) >>> 0) / 4294967296;
-  };
-}
-
-/** 定向混排片段池：ASCII / CJK / emoji（含肤色修饰与 ZWJ 组合）/ SGR / 非 SGR ESC（OSC、CSI-K、裸 ESC）/ 换行。 */
-const FUZZ_PIECES = [
-  "plain", " ", "a", "0", "MiXeD-Case_123",
-  "你好世界", "汉字宽度测试", "日本語テキスト",
-  "😀", "😁😂🤣", "👍🏽", "👨‍👩‍👧‍👦", "é", "é",
-  "\x1b[31m", "\x1b[0m", "\x1b[m", "\x1b[1;32m", "\x1b[38;5;196m",
-  "\x1b[K", "\x1b]0;title\x07", "\x1bZ", "\x1b",
-  "\n", "\r\n",
-];
-
-function genFuzzText(rand: () => number): string {
-  const pieceCount = Math.floor(rand() * 40);
-  let out = "";
-  for (let p = 0; p < pieceCount; p++) {
-    out += FUZZ_PIECES[Math.floor(rand() * FUZZ_PIECES.length)];
-  }
-  return out;
-}
-
-describe("truncLine — legacy 对拍（per-char 参照逐字节等价）", () => {
-  it("定向 fuzz：300 随机 ESC/SGR/CJK/emoji 混排用例（固定 seed 可复现）与旧实现逐字节一致", () => {
-    const rand = mulberry32(20260815);
-    let truncated = 0;
-    for (let i = 0; i < 300; i++) {
-      const text = genFuzzText(rand);
-      const maxWidth = Math.floor(rand() * 61); // 0..60，含 0/1 边界
-      const actual = truncLine(text, maxWidth);
-      expect(
-        actual,
-        `case #${i} text=${JSON.stringify(text)} maxWidth=${maxWidth}`,
-      ).toBe(legacyTruncLine(text, maxWidth));
-      if (actual.endsWith("…")) truncated++;
-    }
-    // 前置有效性：fuzz 确实命中了截断路径（否则对拍只测了「宽度内原样返回」早退分支）
-    expect(truncated).toBeGreaterThan(100);
-  });
-
-  it("手工对抗用例：SGR 重叠 / OSC 内截断 / CSI-K 混排 / 串尾裸 ESC", () => {
-    const cases: Array<[string, number]> = [
-      ["\x1b[31m红\x1b[32m绿\x1b[1;4m加粗下划线尾巴超宽截断点", 9],
-      ["AB\x1b]0;title\x07CDEFGHIJKLMNOP", 10],
-      ["\x1b[31mred\x1b[K\x1b[32mgreen tail past width", 10],
-      ["text almost at width limit\x1b", 22],
-      ["before \x1bZ after bare esc tail", 12],
-      ["\x1b[m你\x1b[0m好\x1b[38;5;196m世👨‍👩‍👧‍👦界".repeat(3), 14],
-    ];
-    for (const [text, maxWidth] of cases) {
-      expect(truncLine(text, maxWidth)).toBe(legacyTruncLine(text, maxWidth));
-    }
-  });
-});
-
-// ============================================================
-// wrapText
-// ============================================================
-describe("wrapText", () => {
-  it("returns text as-is when shorter than width", () => {
-    expect(wrapText("hello", 80)).toEqual(["hello"]);
-  });
-
-  it("wraps long text into multiple lines (no truncation)", () => {
-    const text = "abcdefghij"; // 10 chars
-    const lines = wrapText(text, 4);
-    // 每行最多 4 列，10 字符 → 3 行（4+4+2），不截断不省略号
-    expect(lines.join("")).toBe("abcdefghij");
-    for (const l of lines.slice(0, -1)) {
-      expect(visibleWidth(l)).toBeLessThanOrEqual(4);
-    }
-  });
-
-  it("preserves original newlines as paragraph breaks", () => {
-    const text = "第一行\n第二行";
-    const lines = wrapText(text, 80);
-    expect(lines).toEqual(["第一行", "第二行"]);
-  });
-
-  it("wraps CJK text correctly (2 columns each)", () => {
-    // 你好世界你好世界 = 8 CJK chars = 16 columns, wrap to 4 columns
-    const text = "你好世界你好世界";
-    const lines = wrapText(text, 4);
-    // 每行最多 4 列 = 2 CJK chars, 共 4 行
-    expect(lines).toHaveLength(4);
-    expect(lines.join("")).toBe(text);
-  });
-
-  it("handles width <= 0 by returning original", () => {
-    expect(wrapText("hello", 0)).toEqual(["hello"]);
-  });
-
-  it("preserves empty lines from blank paragraphs", () => {
-    const text = "a\n\nb";
-    expect(wrapText(text, 80)).toEqual(["a", "", "b"]);
-  });
-});
-
-// ============================================================
-// shortId
-// ============================================================
-describe("shortId", () => {
-  it("returns sync id unchanged (run-N already short)", () => {
-    expect(shortId("run-1")).toBe("run-1");
-    expect(shortId("run-42")).toBe("run-42");
-  });
-
-  it("strips timestamp from background id (bg-tag-seq-<ts> → bg-tag-seq)", () => {
-    // 算法回归：多段 id 取前 3 段（SHORT_ID_BG_SEGMENTS）。
-    // 注意：当前 subagent ID 已改为 sa-<uuid>（见下面 sa- 用例），workflow ID 为 wf-<ts>-<rand>，
-    // 实际不再产生 bg- 形态 id；此处保留作为 shortId 算法的多段降级回归（4 段 → 3 段）。
-    expect(shortId("bg-f6f731-10-1719500000000")).toBe("bg-f6f731-10");
-    expect(shortId("bg-abc123-99-1719500123456")).toBe("bg-abc123-99");
-  });
-
-  it("handles pure uuid and wf- runId (regression baseline)", () => {
-    // 纯 UUID 回归（5段 → 取前3段）
-    expect(shortId("550e8400-e29b-41d4-a716-446655440000")).toBe("550e8400-e29b-41d4");
-    // wf- 前缀 runId 回归（3段 → 取前3段=原样）
-    expect(shortId("wf-1719500000000-a1b2c3")).toBe("wf-1719500000000-a1b2c3");
-  });
-
-  it("keeps sa- prefix for subagent id (sa-<uuid> → sa-<uuid 前3段>)", () => {
-    // sa- 前缀 subagent ID（保留前缀 + UUID 前 3 段）
-    expect(shortId("sa-550e8400-e29b-41d4-a716-446655440000")).toBe("sa-550e8400-e29b-41d4");
-  });
-});