This product was not featured by Product Hunt yet. It will not be visible on their landing page and won't be ranked (cannot win product of the day regardless of upvotes).
Spellcast was built with GPT-6 Astra. I used Astra heavily to design and ship the desktop stage: passive bubbles for side thoughts, an editable Canvas for results, and completion notices when a turn ends. That burned through my Astra Pro 20x before launch day. Runtime target today: Codex on Windows (MCP + hooks). Local-first, open source, EN/中文 UI. The clip shows one host path; the core idea is the desktop stage around the agent.
Hi Product Hunt 👋
I built Spellcast because agent sessions kept pulling me back to the terminal. I wanted the agent's side thoughts, its results, and "I'm done" to reach me where I'm actually looking — the desktop — without me babysitting a CLI.
Three things it does:
• Bubbles — the agent's asides float up while it works; star one to keep it, double-click to open it on the board.
• Canvas — the agent writes results onto an editable Canvas as ideas (text, step sequences) that you can keep shaping.
• Completion notices — a card plus a spoken alert the moment a turn ends, for both Codex and Grok Build.
Honest note about the demo: Spellcast was designed around the Codex app/CLI, but my Codex Pro quota ran out right before the deadline, so the video shows the Grok Build integration instead. Everything you see is real and end-to-end, but the Codex-side flow (deep-linking a completion back into the exact session, replies from the Canvas) isn't on camera. I'd have loved to record it with Astra — its computer use is far ahead of what I could recreate with Cursor for driving the demo — but I was out of credits there too.
Everything is open source. Happy to answer anything about the MCP setup, the hook-based completion notices, or the Canvas model.
No comment highlights available yet. Please check back later!
About Spellcast on Product Hunt
“Desktop stage for coding agents: bubbles & Canvas”
Spellcast was submitted on Product Hunt and earned 0 upvotes and 1 comments, placing #57 on the daily leaderboard. Spellcast was built with GPT-6 Astra. I used Astra heavily to design and ship the desktop stage: passive bubbles for side thoughts, an editable Canvas for results, and completion notices when a turn ends. That burned through my Astra Pro 20x before launch day. Runtime target today: Codex on Windows (MCP + hooks). Local-first, open source, EN/中文 UI. The clip shows one host path; the core idea is the desktop stage around the agent.
Spellcast was featured in Productivity (660.9k followers), Developer Tools (519.6k followers), Artificial Intelligence (478.8k followers), GitHub (41.4k followers) and OpenAI Day (38 followers) on Product Hunt. Together, these topics include over 397.4k products, making this a competitive space to launch in.
Who hunted Spellcast?
Spellcast was hunted by Darryl. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
Want to see how Spellcast stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.
Darryl