v1.0.1415 May 2026MAJOR

A web chat that keeps up with the work

Per-chat models and keys, a prompt library on the slash key, attachable skills, MCP servers, side-by-side model compare, fork from any message.

NEW7

Type "/" in the input to drop in a prompt from your library — variables become editable fields right in the chip row.
Attach skills to a chat so the model gets focused domain context before it answers.
Bring your own MCP servers per chat — web search, GitHub, Slack, anything you wire up. The gateway turns them into tools for that session.
Pick the model and API key per chat, not globally. Keep one Claude tab and one GPT-5 tab open side by side.
Compare two models on the same prompt with a split view. Useful when you can't decide between Claude Sonnet 4.6 and Kimi K2.
Fork any message into a new chat to try a different direction without losing the original thread.
Drag, drop, or paste images and files straight into the input. Vision models read them inline.

IMPROVED3

Markdown rendering with syntax-highlighted code blocks and one-click copy on every snippet.
Conversation history restores the original model, API key, and prompt stack — switching chats no longer leaks state.
Export any conversation as Markdown.

RELEASE NOTES

The web chat used to be a thin wrapper around model calls. Now it's an agentic session you can shape per chat.

Per-chat everything

Models, keys, prompts, skills, and MCP servers are all attached to the individual conversation. Open Chat A in Claude Opus 4.7, Chat B in GPT-5, and switching between them just works — no shared selector, no surprise model swap.

The slash key

Type / at the start of the input and your prompt library pops up. Pick one, the template lands in the input, and any {{variables}} show up as fields above the chip. Send when you're ready.

MCP in chat

This is the big one. Add an MCP server to a chat and the gateway creates tools from it on the fly. Anything you can run from the desktop app or our agent collections, you can now run from a one-off web chat. Tool calls render inline.

Compare mode

Click Compare next to the model picker, choose a second model, send a prompt. Two streams, side by side, same prompt. Good for picking a default; better for catching a model that hallucinated.

If you've been waiting for the web chat to catch up to what the desktop app does, this is most of the way there.

Keep up

Most of this shipped itself.

AskCodi maintains AskCodi, so these notes describe work we had to review too.

Start with an idea Get the desktop app