vary-image

name	vary-image
user-invocable	true
description	Make fresh variations of an existing TITLES (titles.xyz) image — same idea, new takes — on a fixed pipeline, no model or prompt to pick. Use when the user says: "variations of this", "more like this one", "some alternates", "different takes on this", "riff on this image", "a few more options from this". Takes an image already on TITLES (its `output_id`) and an optional strength (how far to stray), and returns two new variations on the canvas. Runs on the TITLES MCP — if TITLES tools are missing, connect mcp.titles.xyz/mcp first (see titles-setup). NOT for: the exact same image again (rerun the source generation), a directed change (edit-image / restyle-image), or a new concept (generate-image).

vary-image

New takes on an existing image — same idea, drawn from the image itself. Never ask for the original prompt, never search models.

Get connected

Check the tool list for TITLES tools (names contain titles_). If missing, hand off to the titles-setup skill and stop — never fall back to a non-TITLES tool.

Vary

One call: titles_vary_image with the image's output_id; strength sets how far the takes may stray. Cheap — if the user wants them looser or tighter, re-run at a different strength rather than explaining.

The connected server is authoritative for everything else — exact inputs, model resolution, cost approval (price_confirmation_required → max_price_usd), session/canvas handling, sourcing output_ids, and bringing outside images in (titles_create_upload) all follow the tool's own description and the server instructions, not anything memorized here. titles_help has the current catalog.

Deliver the file

Besides the session_url the server points you at, hand over the actual file via titles_download_asset — host-adaptive: on a shell host (Claude Code / Codex) curl it to disk and give the path; on a chat host (claude.ai / mobile) give the short-lived link to click, re-fetching cheaply if it expires.