Kaiku
Product Hunt Stanislav
Kaiku launched a task tracker and wiki for teams using AI agents. It plugs into existing tracker APIs and logs every agent run’s cost on the issue.
Based on reporting by Product Hunt, Stanislav — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Kaiku is going live as a task tracker and wiki for teams that are handing more work to AI agents. The pitch is pretty specific: instead of asking companies to rip out the tools they already use, Kaiku speaks the API of the tracker they already run, so the MCP servers their agents rely on can keep working unchanged.
That makes the product feel less like a shiny replacement and more like a bridge. It also comes with its own MCP server for Claude Code and Cursor, which should make it easier for agents to show up inside the workflow instead of hovering around it.
The human part is still there, and Kaiku seems to be making that point on purpose. Agents can be called in a comment, they can propose work, but they do not get the final say.
And then there’s the accounting. Every agent run has its cost recorded on the issue it worked on, which is the kind of detail teams will care about once the novelty wears off and the bill arrives.
My take — AI-written commentary, not fact-checked reporting
This is the right instinct: don’t hand the keys to the bot and hope for the best. Let the agent suggest, keep a human in charge, and write the cost down where everyone can see it. That’s less hype, more bookkeeping, which is exactly what AI tooling needs more of.
Read more about this at: Product Hunt