Documentation

Create a new project by clicking New Project in the navigation bar. You need three things:

1. Project Name -- a short, descriptive name for your project.
2. Product Intent -- describe what you want to build in plain language. Be as detailed as you like. The more context you provide, the better the generated requirements and code will be.
3. Regulatory Standards (optional) -- select any industry standards your project must comply with. Compliance documentation will be generated accordingly.

You can also specify a technology stack and additional notes (constraints, performance requirements, integrations).

Once your project is created, click Start Infinity Loop to begin the automated workflow.

At Gate 1, you review every requirement generated by the agents. Each requirement includes:

• A unique code (e.g., REQ-001) for traceability
• Priority level: MUST, SHOULD, or COULD
• Complexity score (1-10) and estimated effort in hours
• Cost projection

For each requirement, you can:

• Accept -- approve it as-is
• Needs Work -- provide specific feedback; the agent will refine the requirement and show you a before/after diff
• Reject -- remove it from the project scope

Bulk actions are available to approve all, reject all, or approve remaining unreviewed requirements. A progress bar tracks your review status.

Change requests let you formally track modifications to your project across cycles. You can create them at Gate 1 or after a project is completed.

Each change request includes a title, description, rationale, impact summary, and lists of affected requirements, artifacts, and test cases. Change requests follow a lifecycle: Proposed → Approved or Rejected → Implemented.

When you reject Gate 2 or start a new cycle, the cycle counter increments. The Cycles tab shows a comparison of requirements, artifacts, test results, and pass rates across all cycles, including a detailed diff view of what changed.

Agile V Studio supports multiple LLM providers. You can configure your preferred provider at two levels:

• User default -- set in Settings; applies to all your projects as a fallback
• Per-project override -- set in Project Settings; takes priority over your default

If you are on a BYOK (Bring Your Own Key) plan, you provide your own API keys and pay the LLM provider directly. Your keys are encrypted before storage and never displayed in plain text. If you are on a Platform plan, we provide the API access and you get a monthly token allowance included in your subscription.

If you run a local LLM (Ollama, LM Studio, vLLM, llama.cpp, or any OpenAI/Anthropic-compatible endpoint), you can use it with Agile V Studio. Your browser acts as the relay between the platform and your local model -- the server never contacts your local network directly.

How it works:

1. Select Custom Local LLM as your provider in Settings or Project Settings.
2. Enter your local endpoint URL (e.g., http://localhost:11434 for Ollama).
3. Choose the API format: OpenAI-compatible (most local LLMs) or Anthropic-compatible.
4. Enter your model name (e.g., llama3.1:70b).
5. Start your Infinity Loop -- a status indicator shows the connection state.

When creating a project, you can select one or more regulatory standards. The Compliance Auditor generates ISO-aligned documentation specific to each standard you choose. Over 30 standards are supported across industries: