Comments API for agents

We've shipped first-class comments support in the API, so AI agents can now read, write, and resolve comments on your drafts.

Check out the API docs here.

What this enables

• Triage on demand. Ask an agent: "Show all my drafts with open comments." Have it resolve the basic copy fixes and surface the editorial decisions worth a human eye.

• An agent editor. You leave a comment: "Hook is too soft." An agent replies in the thread with two rewrite options. You pick one. The edit applies. The thread resolves.

• An agent reviewer. Set up an agent to review each new draft and leave comments on weak hooks, factual claims, voice mismatches, repetition. You could trigger this with a tag like Agent Review.

New endpoints

All endpoints live under /v2/social-sets/{social_set_id}/drafts/{draft_id}/:

• GET /comment-threads: list threads on a draft. Filterable by status and platform.
• POST /comment-threads: create a span-anchored or post-level thread.
• POST /comment-threads/{id}/comments: reply to a thread.
• PATCH /comment-threads/{id}/comments/{comment_id}: edit a comment.
• DELETE /comment-threads/{id}/comments/{comment_id}: delete a comment.
• POST /comment-threads/{id}/resolve: mark a thread resolved.
• DELETE /comment-threads/{id}: delete the whole thread.

Read the updated API docs.

Try it out

• If using our MCP via Claude or ChatGPT, you're already up to date.
• If using our Skill, ask your agents to self-update the typefully skill.