memind/scripts/repair-memory-v2-candidates.mjs

#!/usr/bin/env node
import mysql from 'mysql2/promise';
import { pathToFileURL } from 'node:url';
import { detectFalseStoreCandidate } from '../memory-v2-shadow-audit.mjs';
import { inspectMemoryContentDurability } from '../memory-v2-personal-shadow.mjs';
import { loadMemindEnvFiles } from './memind-runtime-profile.mjs';

function usage() {
  return [
    'Usage: node scripts/repair-memory-v2-candidates.mjs [--apply] [--user-id <id>] [--limit <n>] [--json]',
    '',
    'Finds low-quality accepted/candidate rows and rejects them.',
    'Default mode is dry-run. --apply performs UPDATE status=rejected.',
  ].join('\n');
}

function parseArgs(argv) {
  const args = {
    apply: false,
    userId: '',
    limit: 500,
    json: false,
  };
  for (let i = 2; i < argv.length; i += 1) {
    const arg = argv[i];
    if (arg === '--apply') args.apply = true;
    else if (arg === '--user-id' && argv[i + 1]) args.userId = argv[++i];
    else if (arg === '--limit' && argv[i + 1]) {
      args.limit = Math.max(1, Math.min(5000, Number(argv[++i]) || 500));
    } else if (arg === '--json') args.json = true;
    else if (arg === '--help' || arg === '-h') {
      console.log(usage());
      process.exit(0);
    } else {
      console.error(`Unknown argument: ${arg}`);
      console.error(usage());
      process.exit(2);
    }
  }
  return args;
}

function createPoolFromEnv() {
  loadMemindEnvFiles(process.cwd());
  const poolOptions = { connectionLimit: 3 };
  if (process.env.DATABASE_URL) {
    return mysql.createPool({ uri: process.env.DATABASE_URL, ...poolOptions });
  }
  return mysql.createPool({
    host: process.env.MYSQL_HOST ?? 'localhost',
    port: Number(process.env.MYSQL_PORT ?? 3306),
    user: process.env.MYSQL_USER ?? 'boot',
    password: process.env.MYSQL_PASSWORD ?? '',
    database: process.env.MYSQL_DATABASE ?? 'tkmind',
    ...poolOptions,
  });
}

export function evaluateRepairCandidate(row) {
  const content = String(row.content ?? '');
  const durabilityReason = inspectMemoryContentDurability(content);
  if (durabilityReason) {
    return { repair: true, code: durabilityReason, message: `Non-durable content (${durabilityReason})` };
  }
  const falseStore = detectFalseStoreCandidate(content);
  if (falseStore.suspicious) {
    return { repair: true, code: falseStore.code, message: falseStore.message };
  }
  if (String(row.policy_reason ?? '') === 'decision_signal' && /(?:public\/|page data|durability-)/iu.test(content)) {
    return { repair: true, code: 'decision_signal_task_leak', message: 'Decision signal on agent task content' };
  }
  return { repair: false };
}

async function loadCandidates(pool, { userId, limit }) {
  const clauses = ["status IN ('candidate', 'accepted')"];
  const params = [];
  if (userId) {
    clauses.push('user_id = ?');
    params.push(userId);
  }
  params.push(limit);
  const [rows] = await pool.query(
    `SELECT id, user_id, session_id, memory_type, content, status, policy_reason,
            confidence, importance, created_at, updated_at
     FROM h5_memory_v2_candidates
     WHERE ${clauses.join(' AND ')}
     ORDER BY updated_at DESC
     LIMIT ?`,
    params,
  );
  return rows;
}

async function applyRepairs(pool, repairs, nowMs) {
  let updated = 0;
  for (const item of repairs) {
    const [result] = await pool.query(
      `UPDATE h5_memory_v2_candidates
          SET status = 'rejected',
              reviewed_by = 'system:repair',
              reviewed_at = ?,
              updated_at = ?
        WHERE id = ?
          AND status IN ('candidate', 'accepted')`,
      [nowMs, nowMs, item.id],
    );
    updated += Number(result?.affectedRows ?? 0);
  }
  return updated;
}

function printTextReport(report) {
  console.log(`memory v2 candidate repair: ${report.apply ? 'apply' : 'dry-run'}`);
  console.log(`scanned: ${report.scanned}`);
  console.log(`repairable: ${report.repairable}`);
  if (report.repairs.length > 0) {
    for (const item of report.repairs.slice(0, 20)) {
      console.log(`- [${item.code}] ${item.id} (${item.status}/${item.policyReason}): ${item.contentPreview}`);
    }
    if (report.repairs.length > 20) {
      console.log(`... ${report.repairs.length - 20} more`);
    }
  }
  if (report.apply) {
    console.log(`updated: ${report.updated}`);
  }
}

export async function runMemoryV2CandidateRepair(args, { pool: injectedPool, nowMs = Date.now() } = {}) {
  const ownsPool = !injectedPool;
  const activePool = injectedPool ?? createPoolFromEnv();
  try {
    const rows = await loadCandidates(activePool, args);
    const repairs = rows
      .map((row) => {
        const verdict = evaluateRepairCandidate(row);
        if (!verdict.repair) return null;
        return {
          id: String(row.id),
          userId: String(row.user_id),
          status: String(row.status),
          policyReason: String(row.policy_reason ?? ''),
          code: verdict.code,
          message: verdict.message,
          contentPreview: String(row.content ?? '').slice(0, 120),
        };
      })
      .filter(Boolean);

    const report = {
      ok: true,
      apply: args.apply,
      scanned: rows.length,
      repairable: repairs.length,
      repairs,
      updated: 0,
    };

    if (args.apply && repairs.length > 0) {
      report.updated = await applyRepairs(activePool, repairs, nowMs);
    }
    return report;
  } finally {
    if (ownsPool) {
      await activePool.end();
    }
  }
}

async function main() {
  const args = parseArgs(process.argv);
  try {
    const report = await runMemoryV2CandidateRepair(args);
    if (args.json) {
      console.log(JSON.stringify(report, null, 2));
    } else {
      printTextReport(report);
    }
    process.exit(0);
  } catch (err) {
    const message = err instanceof Error ? err.message : String(err);
    console.error(`repair failed: ${message}`);
    process.exit(1);
  }
}

if (process.argv[1] && import.meta.url === pathToFileURL(process.argv[1]).href) {
  void main();
}