refactor(role): 將角色記憶工具改寫為 Node.js

2026-07-28 12:30:28 +08:00
parent e71f165e9a
commit fec3e1d0cb
8 changed files with 1090 additions and 1125 deletions
@@ -0,0 +1,257 @@
+#!/usr/bin/env node
+// ==============================================================================
+// 用途：角色記憶的 transcript 處理工具。負責 (1) 從 Claude Code／Codex
+//       JSONL 抽出「本輪」對話片段（最後一筆使用者訊息之後的全部內容），
+//       (2) 估算本輪花費時間，(3) 對文字做機密遮蔽（token／密碼／PII），
+//       作為寫入記憶檔前的第二道防線。
+// 更新時間：2026/07/28 12:21:11
+// 相依：Node.js 標準庫。抽取與遮蔽全程僅走 stdin／stdout，本檔不寫任何檔案。
+// ==============================================================================
+
+const fs = require("fs");
+
+const TOOL_RESULT_LIMIT = 200;
+const TOOL_INPUT_LIMIT = 160;
+const TOTAL_LIMIT = 24000;
+
+const REDACT_PATTERNS = [
+  [/[A-Za-z0-9_-]*:[A-Za-z0-9_-]{16,}@/g, "***@"],
+  [/\b[0-9a-f]{40}\b/g, "***"],
+  [/\bgh[pousr]_[A-Za-z0-9_]{16,}\b/g, "***"],
+  [/\bsk-[A-Za-z0-9\-_]{16,}\b/g, "***"],
+  [/\b(token|password|passwd|pwd|secret|api[_-]?key)\b\s*[:=]\s*\S+/gi, "$1=***"],
+  [/Authorization:\s*(token|bearer)\s+\S+/gi, "Authorization: $1 ***"],
+  [/[A-Za-z0-9._%+\-]+@[A-Za-z0-9.\-]+\.[A-Za-z]{2,}/g, "***"],
+  [/\b09\d{2}[-\s]?\d{3}[-\s]?\d{3}\b/g, "***"],
+  [/\b[A-Z][12]\d{8}\b/g, "***"],
+];
+
+function readStdin() {
+  try {
+    return fs.readFileSync(0, "utf8");
+  } catch {
+    return "";
+  }
+}
+
+function redact(text) {
+  let output = String(text || "");
+  for (const [pattern, replacement] of REDACT_PATTERNS) {
+    output = output.replace(pattern, replacement);
+  }
+  return output;
+}
+
+function isObject(value) {
+  return value && typeof value === "object" && !Array.isArray(value);
+}
+
+function isRealUserMessage(entry) {
+  const payload = entry.payload;
+  if (isObject(payload) && entry.type === "event_msg") {
+    return payload.type === "user_message" && Boolean(String(payload.message || "").trim());
+  }
+  if (entry.type !== "user") return false;
+  const content = entry.message?.content;
+  if (typeof content === "string") return Boolean(content.trim());
+  if (Array.isArray(content)) return content.some((block) => isObject(block) && block.type === "text");
+  return false;
+}
+
+function blocks(entry) {
+  const content = entry.message?.content;
+  if (typeof content === "string") return [{ type: "text", text: content }];
+  return Array.isArray(content) ? content : [];
+}
+
+function payloadTextBlocks(content) {
+  if (typeof content === "string") return [content];
+  if (!Array.isArray(content)) return [];
+  const texts = [];
+  for (const block of content) {
+    if (!isObject(block)) continue;
+    if (["input_text", "output_text", "text"].includes(block.type)) {
+      const text = String(block.text || "").trim();
+      if (text) texts.push(text);
+    }
+  }
+  return texts;
+}
+
+function renderCodexPayload(entry) {
+  const payload = entry.payload;
+  if (!isObject(payload)) return [];
+
+  const lines = [];
+  const entryType = entry.type;
+  const payloadType = payload.type;
+
+  if (entryType === "event_msg") {
+    if (payloadType === "user_message") {
+      const message = String(payload.message || "").trim();
+      if (message) lines.push(`[user] ${message}`);
+    } else if (payloadType === "agent_message") {
+      const message = String(payload.message || "").trim();
+      if (message) lines.push(`[assistant:${payload.phase || "assistant"}] ${message}`);
+    }
+    return lines;
+  }
+
+  if (entryType !== "response_item") return lines;
+  if (payloadType === "message") {
+    const role = payload.role || "assistant";
+    if (role === "system" || role === "developer") return lines;
+    for (const text of payloadTextBlocks(payload.content)) {
+      if (role === "user" && text.trimStart().startsWith("<skill>")) continue;
+      if (role === "user" && text.trimStart().startsWith("<environment_context>")) continue;
+      lines.push(`[${role}] ${text}`);
+    }
+  } else if (payloadType === "function_call") {
+    const raw = String(payload.arguments || "").trim().replace(/\n/g, " ");
+    lines.push(`[tool:${payload.name || "?"}] ${raw.slice(0, TOOL_INPUT_LIMIT)}`);
+  } else if (payloadType === "function_call_output") {
+    const raw = String(payload.output || "").trim().replace(/\n/g, " ");
+    if (raw) lines.push(`[result] ${raw.slice(0, TOOL_RESULT_LIMIT)}`);
+  }
+  return lines;
+}
+
+function render(entry) {
+  const codexLines = renderCodexPayload(entry);
+  if (codexLines.length) return codexLines;
+
+  const role = entry.type;
+  const lines = [];
+  for (const block of blocks(entry)) {
+    if (!isObject(block)) continue;
+    if (block.type === "text") {
+      const text = String(block.text || "").trim();
+      if (text) lines.push(`[${role}] ${text}`);
+    } else if (block.type === "tool_use") {
+      const raw = JSON.stringify(block.input || {});
+      lines.push(`[tool:${block.name || "?"}] ${raw.slice(0, TOOL_INPUT_LIMIT)}`);
+    } else if (block.type === "tool_result") {
+      let raw = block.content;
+      if (Array.isArray(raw)) {
+        raw = raw.map((item) => (isObject(item) && item.type === "text" ? item.text || "" : "")).join(" ");
+      }
+      raw = String(raw || "").trim().replace(/\n/g, " ");
+      if (raw) lines.push(`[result] ${raw.slice(0, TOOL_RESULT_LIMIT)}`);
+    }
+  }
+  return lines;
+}
+
+function readEntries(filePath) {
+  let raw;
+  try {
+    raw = fs.readFileSync(filePath, "utf8");
+  } catch {
+    return [];
+  }
+  const entries = [];
+  for (const line of raw.split(/\r?\n/)) {
+    if (!line.trim()) continue;
+    try {
+      entries.push(JSON.parse(line));
+    } catch {}
+  }
+  return entries;
+}
+
+function turnStartIndex(entries) {
+  for (let index = entries.length - 1; index >= 0; index -= 1) {
+    if (isRealUserMessage(entries[index])) return index;
+  }
+  return 0;
+}
+
+function parseTimestamp(value) {
+  if (typeof value !== "string" || !value.trim()) return null;
+  const ms = Date.parse(value.trim());
+  return Number.isNaN(ms) ? null : new Date(ms);
+}
+
+function entryTimestamp(entry) {
+  for (const key of ["timestamp", "created_at", "time"]) {
+    const dt = parseTimestamp(entry[key]);
+    if (dt) return dt;
+  }
+  if (isObject(entry.message)) {
+    for (const key of ["timestamp", "created_at", "time"]) {
+      const dt = parseTimestamp(entry.message[key]);
+      if (dt) return dt;
+    }
+  }
+  return null;
+}
+
+function formatDuration(seconds) {
+  if (seconds < 0) return "未判定";
+  const minutes = Math.round(seconds / 60);
+  if (minutes <= 0) return "1 分鐘內";
+  const hours = Math.floor(minutes / 60);
+  const mins = minutes % 60;
+  if (hours && mins) return `${hours} 小時 ${mins} 分鐘`;
+  if (hours) return `${hours} 小時`;
+  return `${mins} 分鐘`;
+}
+
+function turnDuration(filePath) {
+  const entries = readEntries(filePath);
+  if (!entries.length) return "未判定";
+  const start = turnStartIndex(entries);
+  const stamps = entries.slice(start).map(entryTimestamp).filter(Boolean);
+  if (stamps.length < 2) return "未判定";
+  const min = Math.min(...stamps.map((dt) => dt.getTime()));
+  const max = Math.max(...stamps.map((dt) => dt.getTime()));
+  return formatDuration((max - min) / 1000);
+}
+
+function extractTurn(filePath) {
+  const entries = readEntries(filePath);
+  if (!entries.length) return "";
+  const start = turnStartIndex(entries);
+  const lines = [];
+  for (const entry of entries.slice(start)) lines.push(...render(entry));
+  let text = lines.join("\n").trim();
+  if (text.length > TOTAL_LIMIT) {
+    const half = Math.floor(TOTAL_LIMIT / 2);
+    text = `${text.slice(0, half)}\n…（中段省略）…\n${text.slice(-half)}`;
+  }
+  return text;
+}
+
+const USAGE = `用法：transcript.js <子命令> [參數]
+
+  extract <transcript 路徑>   抽出本輪內容並遮蔽機密後輸出到 stdout
+  duration <transcript 路徑>  估算本輪花費時間，無法判定時輸出「未判定」
+  redact                      自 stdin 讀取文字，遮蔽機密後輸出到 stdout
+`;
+
+function main(argv) {
+  if (!argv.length || argv[0] === "-h" || argv[0] === "--help") {
+    process.stdout.write(USAGE);
+    return 0;
+  }
+  if (argv[0] === "extract") {
+    if (argv.length < 2) return 2;
+    const text = extractTurn(argv[1]);
+    if (!text) return 1;
+    process.stdout.write(redact(text));
+    return 0;
+  }
+  if (argv[0] === "duration") {
+    if (argv.length < 2) return 2;
+    process.stdout.write(turnDuration(argv[1]));
+    return 0;
+  }
+  if (argv[0] === "redact") {
+    process.stdout.write(redact(readStdin()));
+    return 0;
+  }
+  process.stdout.write(USAGE);
+  return 2;
+}
+
+process.exit(main(process.argv.slice(2)));