Solutions
For studios and teams that sell agent-built development
Keep the requirements of every client project as a model with checks, so that scope, code and evidence stay together.
Who this is for
In your terms
You deliver software to clients and use AI agents to write much of the code. The code is quick to produce. What is slow is agreeing what to build, and showing the client afterwards that it was built.
You need the scope of each project written down in a form that an agent can read and a client can check.
Today
What you get today
- Each project keeps its requirements as files in its own repository, with addresses, statuses and acceptance criteria inside.
- Requirements are linked to code with @implements markers, and spec-editor validate checks the links.
- Sources from Jira, GitHub Issues, Slack, Discord and Telegram arrive as sources.
- Export of SRS in the style of IEEE 830, as Markdown or styled HTML, to hand to a client.
- External agents such as Claude Code and Cursor get exactly the elements they need over MCP.
- A delivery team of analysts, a coder, a tester and devops, with review stages decided by evidence.Pro
- Cost attribution per source and role, and budget guards.Pro
- Several client projects in one installation.Pro
- A user manual and a demo video for the client, made from the scenarios of the model.Pro
- Work runs on your own subscriptions; we do not sell tokens.
Coming soon
What is coming
- Seats for a studio's clients coming soon
Next step
Start with the open core
Start with the open core on one project. Pro is in early access; you can request it.