Chat

Chat lets you build naming presets and rules by describing what you need — no fields to configure, no conditions to assemble. Tell NameQuick "invoices should be renamed to date, vendor, and amount, then moved into Invoices by year" and it drafts the filename format or rule for you. You review exactly what changes before anything is saved.

---

What you can build in chat#

Filename formats — describe what to extract and how to format it. Chat adds and reorders fields, sets the filename pattern and separator, and adds cleanup rules. If something's ambiguous, it asks one quick question instead of guessing.

Written instructions — refine free-form naming instructions conversationally, then save the result as a naming preset.

Rules — describe an automation ("move screenshots to a Screenshots folder and tag them") and chat stages a complete rule: conditions, phase, and actions like moving, tagging, labeling, or archiving. You can also open an existing rule in chat and ask for changes.

How it works#

1. Describe what you want. Type it, or pick a starter. In the Naming Presets workspace, chat lives in an assistant pane beside the editor; in Rules, an Editor | Chat toggle switches between hand-editing and conversation.
2. Review the draft. Chat never edits silently — every proposal arrives as a draft card showing exactly what was added, removed, and kept. The draft itself is still hand-editable: tweak a field or separator directly on the card.
3. Test it on a real file. Attach a sample and NameQuick runs the draft against it — you see the generated filename and extracted values for a filename format, or whether the sample matches your rule, before anything is saved.
4. Apply it — or keep talking. Save when it's right. A follow-up message refines the draft; Undo last turn steps back; canceling or errors put your message right back in the composer.

Works with your provider#

Chat runs on whatever AI setup you already use: Managed, your own API key (OpenAI, Anthropic, Gemini, OpenRouter — never silently rerouted), or a local model via Ollama, LM Studio, or MLX. NameQuick adapts how it talks to each model behind the scenes, so weaker local models build drafts step-by-step instead of failing — though for filename-format authoring, cloud models give noticeably better results.

Built to be safe#

Drafts are staged, never auto-applied. Invalid model output gets one automatic repair attempt, then a plain explanation and a retry button. Everything chat creates lands in the same editors you already know — so anything it builds, you can inspect and change by hand.

---

Chat is coming soon. Want it sooner? Tell us on the roadmap. Until then, the Naming Presets and Rules editors do everything chat does — just by hand.