⚠️ Additions carry a URL but this tracker has no URL…

Error message

⚠️  Additions carry a URL but this tracker has no URL column — URL dedup is INACTIVE (fuzzy fallback). Add a `URL` header column to enable it.

What it means

URL-based dedup (Pass 0) needs a URL column in the tracker header to store and compare normalized posting URLs. When additions carry URLs but the tracker lacks the column, merge-tracker.mjs cannot engage the deterministic tier and dedup silently degrades to fuzzy company+role matching — so it warns once (warnedNoUrlCol) instead of degrading invisibly.

Solutions

Add a 'URL' header column to data/applications.md (empty for old rows is fine) and re-run the merge.
Run node merge-tracker.mjs --backfill-urls to backfill URLs from existing tracker reports, then merge.
Accept fuzzy dedup if your pipeline has no duplicate URLs — the warning is informational.
Standardize batch workers to always emit the url column so dedup stays deterministic.

Example fix

# before (tracker header)
| # | Date | Company | Role | Score | Status | PDF | Report | Notes |
# after
| # | Date | Company | Role | Score | Status | PDF | Report | Notes | URL |

Defensive patterns

Strategy: validation

Validate before calling

const trackerHeader = fs.readFileSync('data/applications.md','utf8').split('\n')[0];
if (addition.url && !/\bURL\b/i.test(trackerHeader)) console.warn('Add a `URL` header column to data/applications.md to enable URL dedup');

Prevention

Include a URL column in the tracker header from day one.
Emit the url field in every batch TSV so deterministic Pass 0 dedup engages.
Run node merge-tracker.mjs --backfill-urls after adding the column to hydrate old rows.
Re-check the header after any manual tracker edit.

When it happens

Trigger: Any merge where addition.url is set (the TSV's url column or a trailing http(s):// field) and COLMAP.url is null — i.e. data/applications.md has no 'URL' header column.

Common situations: Tracker created before the URL column feature; a hand-maintained minimal tracker; an old template lacking the column — same-position reposts then dedup poorly.

Related errors

AI-assisted analysis of santifer/career-ops@aac998c7ed (2026-09-16). Data as JSON: /api/errors/dc04c8c68f9fad72. Report an issue: GitHub.

Appendix: source

Thrown at merge-tracker.mjs:1378

  const content = readFileSync(join(ADDITIONS_DIR, file), 'utf-8');
  const addition = parseTsvContent(content, file);
  if (!addition) { skipped++; continue; }

  // A via= tag can only be stored if the tracker has a Via column — warn
  // instead of dropping the channel silently (#1596). Clear the value too:
  // existing rows parse with via='' on this layout, so a set addition.via would
  // make the cross-channel duplicate guard see a channel mismatch and add a
  // second ? row instead of updating the same-agency re-blast.
  if (addition.via && COLMAP.via == null) {
    console.warn(`⚠️  ${file}: carries via=${addition.via} but the tracker has no Via column — value dropped. Add it with: node merge-tracker.mjs --migrate-via`);
    addition.via = '';
  }

  // If additions carry a URL but the tracker has no URL
  // column, deterministic Pass 0 cannot engage and dedup silently falls back to
  // the fuzzy tiers. Warn once rather than degrade invisibly.
  if (addition.url && COLMAP.url == null && !warnedNoUrlCol) {
    console.warn('⚠️  Additions carry a URL but this tracker has no URL column — URL dedup is INACTIVE (fuzzy fallback). Add a `URL` header column to enable it.');
    warnedNoUrlCol = true;
  }

  // Normalize the report link to be relative to the tracker file's directory.
  // The TSV convention carries a root-relative `reports/...` link; rewrite it
  // so it resolves correctly when clicked from applications.md (see #760).
  addition.report = normalizeReportLink(addition.report);

  // Derive the key from the linked report when the TSV did not carry one, so a
  // row added through the normal nine-column flow is born WITH its key. This
  // must run before the dedup below reads addition.url — deriving it afterwards
  // would populate the column while leaving Pass 0 nothing to match on, which is
  // the original bug with extra steps.
  if (!addition.url) addition.url = resolveReportUrl(addition.report).url;

  // Check for duplicate by:
  // 0. Exact normalized posting URL (deterministic, authoritative)
  // 1. Exact report number match

View on GitHub (pinned to aac998c7ed)