Most documentation tools give you somewhere to write. Thally also keeps what you wrote true. When a product pull request merges, it works out exactly what changed, traces the impact across your docs, website, and support platform, then opens evidence-backed PRs for your team to review. It is also a complete, AI-native docs platform with Git-based content, deploy previews, search, analytics, custom domains, self-hosting and a remote MCP server.
Hi Product Hunt, I’m Ekene, founder of Thally.
For more than seven years, I worked in developer experience at LI.FI, Netlify, and Flutterwave.
At every company, launches created the same job. I had to find out what actually changed, who it affected, and which guides, API references, examples, website pages, support articles, and migration guides were now out of date.
The writing was rarely ever the hard part. It was always the manual detective work that followed the launches. Important context was scattered across product briefs, tickets, pull requests, code, tests, and conversations with product managers and engineers.
As teams shipped faster, the product and its documentation eventually drifted apart.
A guide would describe an older workflow. An example would use a deprecated parameter. A newly released feature would be missing entirely. Different places would tell customers different versions of the same story.
AI-assisted development makes this problem more urgent. Developers increasingly rely on coding assistants to understand products and build integrations. When those assistants learn from outdated documentation, they generate outdated code and recommend deprecated workflows.
No documentation team can solve that by simply writing faster.
That is why I’m building Thally.
Thally is the product knowledge layer for software teams. It monitors the repositories where products are built, works out what a merged change means, traces its impact across your docs, website, help center, changelog, and other connected content, then opens evidence-backed pull requests for the affected pages.
Your team can see why each update was proposed, review the evidence, and approve what ships. When nothing needs to change, Thally says so instead of inventing work.
Thally is also a complete documentation platform. Teams write in MDX, then Thally serves the same source as polished HTML for people, Markdown and structured JSON for tools, plus llms.txt and a live MCP endpoint for AI agents.
Your content remains in your Git repository. You can use Thally Cloud or self-host the open-source engine.
Thally is currently in beta, so there are still rough edges, particularly as we encounter more repository structures and publishing workflows. That is exactly the feedback I’m hoping to get from this launch.
If you have ever discovered that a launch quietly made part of your documentation wrong, I would love to hear how your team handles it today.
You can try Thally at https://thally.io/