Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/LIDR-academy/lidr-specboot/llms.txt

Use this file to discover all available pages before exploring further.

MCP (Model Context Protocol) servers extend your AI copilot’s capabilities by giving it direct, structured access to external tools — without copy-paste or manual context injection. In the context of Specboot’s OpenSpec workflow, two MCP integrations are specifically referenced in the skill definitions: a Jira MCP for reading ticket details during story enrichment, and a Playwright MCP for running browser-based end-to-end checks during verification. Both integrations slot into the existing workflow naturally, reducing friction and improving the quality of what the agent produces.
Both MCP integrations are optional. The full OpenSpec workflow — /enrich-us, /ff, /apply, /verify, /adversarial-review, /archive, /commit — works without either integration. If you don’t use Jira, or you don’t want automated E2E testing, you can skip these entirely. You can also replace them with equivalent tools by updating the relevant skill files in ai-specs/skills/.

​
Supported Integrations

Jira MCP

Lets the agent read Jira ticket details directly during /enrich-us, eliminating manual copy-paste of story content.

Playwright MCP

Lets the agent run browser-based E2E checks during /verify for user-facing features, providing live validation against a running app.

​
Jira MCP

The Jira MCP is used by the enrich-us skill when you pass a Jira ticket ID (e.g., SCRUM-10) instead of the full ticket text. Without it, the skill operates in direct input mode — you paste the ticket content into the chat and the skill works on that. With it, the skill fetches the ticket directly from your Jira instance, enriches the story, and can write the enhanced version back to the ticket automatically.

​
When it is used

The enrich-us skill determines the input source automatically:
  • Direct input mode (default): If you provide full ticket text in the chat, Jira MCP is not used. The skill processes the text you supply.
  • Jira mode (optional): If you provide a Jira ticket ID (e.g., SCRUM-10) or ask the agent to use “the ticket in progress” or similar, the skill uses Jira MCP to fetch the ticket details.
In Jira mode, after enriching the story, the skill can also update the Jira ticket by appending the enhanced content after the original — with clear [original] and [enhanced] sections. If the ticket status is To refine, it moves it to Pending refinement validation.

​
How to configure (Cursor)

  1. Open Cursor Settings and navigate to the MCP section.
  2. Add a new MCP server entry for your Jira instance.
  3. Provide the required credentials — typically a Jira API token (generated from your Atlassian account settings), your Jira base URL, and your user email.
  4. Save and restart Cursor to activate the connection.

​
How to configure (other agents/IDEs)

Follow your tool’s MCP installation documentation. The general pattern is the same: register the Jira MCP server endpoint and supply your API token as a credential. Consult your IDE’s MCP setup guide for the exact configuration format.

​
Playwright MCP

The Playwright MCP is used during the /verify step for features that have a user-facing browser interface. Rather than relying solely on unit and integration tests, the verify workflow can use Playwright to spin up a browser, navigate the application, and validate acceptance criteria against a running instance.

​
When it is used

The Playwright MCP is invoked automatically during /verify when the feature being verified includes browser-based acceptance scenarios. For purely backend or non-UI features, the verify step proceeds without it. Testing reports from Playwright runs are stored in the OpenSpec change directory alongside the other verification artifacts.

​
How to configure (Cursor)

  1. Open Cursor Settings and navigate to the MCP section.
  2. Add the Playwright MCP server. This typically involves specifying the server package or local binary path and any launch options (e.g., headless mode, browser choice).
  3. Save and restart Cursor.

​
How to configure (other agents/IDEs)

Follow your tool’s MCP installation documentation. Playwright MCP implementations vary — some run as a local server process, others as a CLI-driven adapter. Refer to the MCP provider’s documentation for setup instructions specific to your environment.

​
Using /enrich-us Without Jira

You never need Jira to use /enrich-us. In direct input mode, simply paste the full ticket text — title, description, acceptance criteria, and any context — directly into the chat when you invoke the skill. The enrich-us skill will analyze and enhance it exactly as it would with a Jira-fetched ticket, without requiring any MCP connection.This is the default behavior and works in every environment. Use Jira mode only when you want the convenience of fetching tickets by ID and writing enriched content back to Jira automatically.
Example: invoking enrich-us in direct input mode — 
/enrich-us

## User Story: Add candidate filters to position list

As a recruiter, I want to filter candidates by status so that I can focus on
the ones that need action.

Acceptance criteria:
- Filter by: pending, reviewed, shortlisted, rejected
- Filter state persists on page reload
- Filter count shown in UI
The skill will validate completeness, identify missing technical detail (endpoints, files to modify, non-functional requirements), and return an ## Original / ## Enhanced breakdown.

​
Replacing MCPs with Alternative Tools

If your team uses a project management tool other than Jira, or a different browser automation framework instead of Playwright, you can adapt the workflow to use them:
  1. Open the relevant skill file in ai-specs/skills/ — enrich-us/SKILL.md for the ticket enrichment workflow, and the verify-related skill for E2E testing.
  2. Update the instructions to reference your tool’s MCP server or CLI interface instead of Jira/Playwright.
  3. Update any credential or configuration references to match your tool’s authentication model.
  4. Verify that the symlinks in .claude/skills/ and .cursor/skills/ still point to the updated canonical files in ai-specs/skills/.
Since skills are plain Markdown files with structured instructions, swapping out the underlying tool is a documentation change — no code changes required.

OpenSpec Workflow

See where Jira MCP and Playwright MCP fit into the full /enrich-us → /commit sequence.

enrich-us Skill

Full reference for the enrich-us skill — both direct input and Jira mode.

Skills Overview

Browse all available skills and learn how to write your own.

Customization

Adapt agents and skills to your team’s tools, including non-Jira workflows.

Build docs developers (and LLMs) love

Get started for freeTalk to us