Renaming the new-doc Skill to md-artifact

---json { "title": "Renaming the new-doc Skill to md-artifact", "slug": "renaming-new-doc-to-md-artifact", "date": "2026-03-17T02:50:52.460Z", "description": "Why we renamed the new-doc skill to md-artifact and tightened up its trigger description.", "type": "post", "author": [ "Jay Griffin", "Claude Sonnet 3.7" ], "authorshipNote": "Written collaboratively with Claude Sonnet 3.7", "tags": [ "ai", "skills", "cline", "agents", "postmaster" ], "draft": false }

The new-doc Cline skill got a rename today. It's now md-artifact.

Why

The original name and description were too generic. "Create a new markdown document summarizing a conversation or topic" could match almost anything - write a file, save some notes, document some code. The agent was at risk of firing this skill when you just wanted a quick file written somewhere, not a full structured post routed through Postmaster.

The fix was to make the trigger more intentional. md-artifact borrows from Anthropic's "artifact" concept - a deliberate, structured output meant to be reviewed and dispatched, not just a file that gets written and forgotten. The description now explicitly says not to use it for general file writing or code tasks.

What Changed

• Skill directory renamed from new-doc/ to md-artifact/
• Skill name and description updated in SKILL.md
• Script path updated to reflect the new directory
• Port reference fixed (was pointing to localhost:3000, Postmaster runs on 3333)
• Removed the instruction to check if the drafts/ directory exists - it's part of the Postmaster repo, it's always there
• Clarified the body-writing step: read the draft to get the script-generated frontmatter, don't rewrite it, just append the body

The Naming Logic

The name md-artifact is scoped intentionally. If other skills produce artifacts in the future (JSON artifacts, code artifacts, whatever), they'd be json-artifact, code-artifact, etc. The md- prefix keeps this one clearly in its lane.