@zhushanwen/pi-subagent-workflow 8.5.0 → 8.7.0

package/workflows/parallel.js

@@ -1,148 +0,0 @@
-// parallel.js — 并行多视角分析（通用 subagent 编排）
-//
-// 模式：N 个 agent 从不同角度并行分析同一目标 → 汇总聚合。
-// 适用于需要多维度评估（安全/性能/可维护性等）的场景。
-//
-// 用法：
-//   workflow run parallel --args target="src/auth/login.ts"
-//   workflow run parallel --args target="..." --args 'perspectives=["security","readability"]'
-//
-// ⚠️ 分层配额规则（来源：ADR-030 决策 3）：
-//   - 全局并发上限 maxConcurrent = 6
-//   - parallel() 内的 agent() 调用共享配额池，超出自动排队（不报错）
-//   - 本脚本默认 3 视角并行，配额充足
-//
-// ⚠️ lintScript 约束（本脚本已遵守）：含 parallel() 入口，禁止 bare IIFE
-
-/* @pi-meta
-name: parallel
-description: 通用编排：多视角并行分析同一目标，再聚合汇总
-phases: [parallel-analyze, aggregate]
-parameters:
-  type: object
-  properties:
-    target: { type: string }
-    perspectives: { type: array, items: { type: string } }
-    agents: { type: string }
-  required: [target]
-usage: |
-  ## 使用说明
-  - 多视角并行分析后聚合；perspectives 缺省用 3 个默认视角
-  - agents：逗号分隔的 agent .md 绝对路径；1 个应用于所有视角，N 个一一对应视角数
-  - 示例：workflow run parallel --args target="<分析目标>" agents="/path/analyst.md"
-*/
-
-// ── 入参（$ARGS）──────────────────────────────────────────────────
-const target = $ARGS.target;
-if (!target) {
-  throw new Error("parallel 缺少必需参数 target。用法：workflow run parallel --args target=\"<分析目标>\"");
-}
-const perspectives = Array.isArray($ARGS.perspectives) && $ARGS.perspectives.length > 0
-  ? $ARGS.perspectives
-  : ["security", "performance", "maintainability"];
-if (perspectives.some((p) => typeof p !== "string")) {
-  throw new Error("parallel 参数 perspectives 必须是字符串数组，实际含非字符串元素");
-}
-
-// S4：agents 参数 = 逗号分隔的 agentRef 路径数组；1 个 = 所有视角，N 个 = 一一对应
-// worker 沙箱为 eval 模式：require 相对路径以 cwd 为基准（非脚本目录），
-// 必须用 workerData.scriptPath 锚定脚本目录（review-fix-loop 同模式）。
-const SCRIPT_DIR = workerData && workerData.scriptPath
-  ? require("path").dirname(workerData.scriptPath)
-  : process.cwd();
-const { parseAgentRefs, agentRefAt } = require(SCRIPT_DIR + "/_shared/agent-refs.cjs");
-const agentRefs = parseAgentRefs($ARGS.agents);
-const agentFor = (i) => {
-  if (agentRefs.length === 1) return { agent: agentRefs[0] };
-  const ref = agentRefAt(agentRefs, i);
-  return ref ? { agent: ref } : {};
-};
-
-log("parallel 开始，target=" + target + " perspectives=" + JSON.stringify(perspectives) + (agentRefs.length ? " agents=" + agentRefs.join(",") : ""));
-
-let currentPhase = "init";
-let outcome;
-
-try {
-  // ── 段 1：parallel-analyze（多视角并行分析）──────────────────────
-  phase("parallel-analyze");
-  currentPhase = "parallel-analyze";
-
-  // parallel() 接受 Promise 数组；agent() 返回 Promise。allSettled 语义。
-  const perPerspectiveRaw = await parallel(
-    perspectives.map((p, i) =>
-      agent({
-        prompt:
-          "从「" + p + "」角度分析以下目标，给出评分和发现的问题：\n\n" + target,
-        schema: {
-          type: "object",
-          properties: {
-            perspective: { type: "string", description: "视角名称" },
-            score: { type: "number", description: "0-10 评分" },
-            findings: {
-              type: "array",
-              items: { type: "string" },
-              description: "发现的问题",
-            },
-          },
-          required: ["perspective", "score", "findings"],
-        },
-        description: "parallel-" + p,
-        ...agentFor(i),
-      })
-    ),
-  );
-
-  // 收集结果，标记成功/失败
-  const perPerspective = [];
-  let failedCount = 0;
-  for (let i = 0; i < perPerspectiveRaw.length; i++) {
-    const r = perPerspectiveRaw[i];
-    if (!r || r.status === "failed" || r.error) {
-      perPerspective.push({
-        perspective: perspectives[i],
-        status: "failed",
-        error: r ? (r.error || "agent 返回 failed 状态") : "agent 无返回",
-      });
-      failedCount++;
-    } else {
-      perPerspective.push({
-        perspective: perspectives[i],
-        status: "ok",
-        score: typeof r.score === "number" ? r.score : undefined,
-        findings: Array.isArray(r.findings) ? r.findings : [],
-      });
-    }
-  }
-  if (failedCount === perspectives.length) {
-    throw new Error("全部视角分析失败（" + failedCount + "/" + perspectives.length + "）");
-  }
-  log("parallel-analyze 完成：ok=" + (perspectives.length - failedCount) + " failed=" + failedCount);
-
-  // ── 段 2：aggregate（汇总多视角结果）────────────────────────────
-  phase("aggregate");
-  currentPhase = "aggregate";
-
-  // 纯代码合并：拼接各视角发现的问题（不调用 LLM）
-  const aggregateResult = perPerspective
-    .map((p) => "[" + (p.perspective || "?") + "] " + (p.findings ? p.findings.join("; ") : "(no findings)"))
-    .join("\n");
-
-  outcome = {
-    status: failedCount > 0 ? "partial" : "ok",
-    phases_run: ["parallel-analyze", "aggregate"],
-    perspectives_analyzed: perspectives.length,
-    per_perspective: perPerspective,
-    aggregate: aggregateResult,
-    message: "parallel 完成：" + perspectives.length + " 视角（失败 " + failedCount + "）→ 聚合",
-  };
-} catch (err) {
-  outcome = {
-    status: "error",
-    phase: currentPhase,
-    error: err && err.message ? err.message : String(err),
-    message: "parallel 在 " + currentPhase + " 段失败",
-  };
-}
-
-return outcome;