Illia Shapoval
Free Lesson

Set Up a Workspace for Agentic Requirements Authoring

30 min
Sep 29, 2026 11:00 AM

By continuing, you agree to Maven's Terms and Privacy Policy.

What you'll learn

Control what the agent reads before it writes a spec

Add and reference context files, project rules and skills through the file system. See the draft change with them.

Set up an authoring workspace in VS Code with no code

Install VS Code, a markdown editor and an agent plugin for your LLM subscription. Write specs where the agent works.

Publish specs to your team and tools from the workspace

Push to git for developers and UX, send to Jira or Confluence through MCP. One source, no copy-paste

Why this topic matters

In a chat window you paste context, get a draft and paste it somewhere else. In an IDE the LLM works as an agent: it reads your context base, follows your rules and skills, writes the spec file and hands it to git or Jira through MCP. The draft improves as the agent reads the context base. The setup takes one evening and no code. This lesson walks BAs and POs through it, from the minimal kit to the agent doing the authoring with free tools.

You'll learn from

Illia Shapoval

Illia Shapoval

PM & BA, 15 yrs in Banking, Aviation, MedTech | AI practitioner

Lufthansa Systems
Backbase
Luxoft
UKRSIBBANK BNP Paribas Group
See all products from Illia
Get free access