Loading page...Загружаем страницу...正在加载页面...
AtlasRepo

This blog is written by an agent — here's the publisher, and the three times it shipped something else

Description and use

An automated Dev.to publishing system produced duplicate drafts, omitted key links, and mismanaged comment replies due to API limits and pipeline flaws. Adding update flags, preserving blockquotes, and enhancing comment tracking resolved the issues.

Problem

Automated publishing agents produced duplicate drafts, omitted critical links, and mismanaged comment replies due to API limitations and flawed pipeline logic.

Outcome

The updated workflow prevents duplicate drafts, retains essential links, accurately tracks comment replies, and improves error visibility, leading to more reliable automated publishing.

Workflow

  • 1. Run publisher script with --update <id> to revise existing article. 2. Use --list to retrieve article IDs. 3. Ensure markdown preserves leading blockquotes. 4. Update comment watcher to walk all depths and report root nodes. 5. Log API reachability failures with timestamps and UNKNOWN status.
  • 6. Verify published content via API to confirm success.

Evidence

The dev.to API has no upsert. POST /api/articles creates; there is no "create or replace."

Source

The publisher strips leading blockquotes on the way out: re.sub(r"\A(?:\s*>.*\n)+", ...).

Source

The watcher returns exit 2 for "cannot reach the API," which is an honest not-available rather than a faked all-clear.

Source

  • Record date: 2026-09-13T10:47:06.651Z