linkedin-studio/scripts/brain/src/cli.ts

#!/usr/bin/env node
/**
 * CLI for the second-brain foundation + ingest (SB-S0 / SB-S1).
 *
 *   node --import tsx src/cli.ts init
 *   node --import tsx src/cli.ts ingest --file <path> [--source <s>] [--date <YYYY-MM-DD>]
 *   node --import tsx src/cli.ts ingest --scan-inbox [--source <s>]
 *   node --import tsx src/cli.ts published list [--json]
 *
 * `init` scaffolds the brain/ + ingest/ tree (idempotent). `ingest` captures the
 * user's published posts into ingest/published/ tagged provenance=published — the
 * gold signal the voice/profile-learning surface learns from (never ai-draft).
 *
 * Exit code: 0 on success, 2 on usage error.
 */

import { readFileSync } from "node:fs";

import { dataRoot } from "./dataRoot.js";
import { ingestText, listPublished, scanInbox } from "./ingest.js";
import { initBrain } from "./scaffold.js";

/** Parse `--key value` / boolean `--flag` args (the specifics-bank CLI idiom). */
function parseFlags(args: string[]): Record<string, string> {
  const out: Record<string, string> = {};
  for (let i = 0; i < args.length; i++) {
    const a = args[i];
    if (a.startsWith("--")) {
      const key = a.slice(2);
      const next = args[i + 1];
      if (next === undefined || next.startsWith("--")) {
        out[key] = "true";
      } else {
        out[key] = next;
        i++;
      }
    }
  }
  return out;
}

function today(): string {
  return new Date().toISOString().slice(0, 10);
}

function usage(msg: string): never {
  console.error(`error: ${msg}`);
  console.error(
    "usage:\n" +
      "  init\n" +
      "  ingest --file <path> [--source <s>] [--date <YYYY-MM-DD>]\n" +
      "  ingest --scan-inbox [--source <s>]\n" +
      "  published list [--json]",
  );
  process.exit(2);
}

function runInit(): void {
  const { created, skipped } = initBrain();
  console.log(`Brain root: ${dataRoot("brain")}`);
  if (created.length > 0) {
    console.log(`Created (${created.length}):`);
    for (const c of created) console.log(`  + ${c}`);
  }
  if (skipped.length > 0) {
    console.log(`Skipped — already present (${skipped.length}):`);
    for (const s of skipped) console.log(`  · ${s}`);
  }
  if (created.length === 0) console.log("Already initialised — nothing to do.");
}

function runIngest(flags: Record<string, string>): void {
  if (flags["scan-inbox"] === "true") {
    const res = scanInbox({ captured_at: today(), source: flags.source });
    console.log(
      `Scanned inbox → published: ${res.processed.length} new, ${res.skipped.length} already published`,
    );
    return;
  }
  const file = flags.file;
  if (!file || file === "true") usage("ingest needs --file <path> or --scan-inbox");
  const body = readFileSync(file, "utf8");
  const res = ingestText({
    body,
    captured_at: today(),
    source: flags.source,
    published_date: flags.date === "true" ? undefined : flags.date,
  });
  if (res.written && res.collision) {
    console.log(`Collision (different body, same id) → wrote ${res.path}`);
  } else if (res.written) {
    console.log(`Wrote ingest/published/${res.record.id}.md`);
  } else {
    console.log(`Duplicate — already published: ${res.record.id}`);
  }
}

function runPublished(rest: string[], flags: Record<string, string>): void {
  if (rest[0] !== "list") usage("published needs: list");
  const { records, skipped } = listPublished();
  if (flags.json === "true") {
    console.log(JSON.stringify({ records, skipped }, null, 2));
    return;
  }
  const tail = skipped > 0 ? ` (${skipped} unreadable skipped)` : "";
  console.log(`Published gold corpus: ${records.length} record(s)${tail}`);
  for (const r of records) {
    console.log(`  · ${r.id} · ${r.provenance} · ${r.published_date} · ${r.firstLine}`);
  }
}

function main(): void {
  const [command, ...rest] = process.argv.slice(2);
  const flags = parseFlags(rest);

  if (command === "init") return runInit();
  if (command === "ingest") return runIngest(flags);
  if (command === "published") return runPublished(rest, flags);

  usage(command ? `unknown command: ${command}` : "no command given");
}

main();