From mida-skills
Use when building a new feature from a Jira ticket — orchestrates Jira MCP → Figma MCP → Shopify Dev MCP → repo skill into a single implementation workflow.
How this skill is triggered — by the user, by Claude, or both
Slash command
/mida-skills:mida-feature-workflowThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this skill when given a Jira ticket (e.g. `MAH-123`) to implement. It sequences all available MCPs and repo skills to go from task → working code.
Use this skill when given a Jira ticket (e.g. MAH-123) to implement. It sequences all available MCPs and repo skills to go from task → working code.
MAH-xxx, MIDA-xxx, etc.)Use Atlassian Rovo MCP (cloud OAuth, no API token needed):
mcp__claude_ai_Atlassian_Rovo__getJiraIssue: { issueIdOrKey: "<ISSUE_KEY>" }
If not authenticated: tell user to run /mcp → select claude.ai Atlassian Rovo → authenticate.
Extract:
mida-cms, mida-api)figma.com/file/... or figma.com/design/...)If the ticket has subtasks, read each one too.
Use the figma-developer-mcp with the URL from Step 1:
get_figma_data: { fileKey, nodeId? }
Extract from the design:
<BlockStack> / <InlineStack> + gap tokentone props (critical, warning, success, info)<Text variant="..."> valuesIf there are multiple frames (desktop / mobile / states), read the primary one first.
Use the shopify-dev-mcp to confirm correct component usage:
search_docs_chunks: "component name + usage context"
validate_component_codeblocks: <generated JSX snippet>
Use this to:
mida-skills:mida-cms)Based on which repo(s) the Jira ticket affects, invoke the matching skill before writing any code:
| Affected area | Skill to invoke |
|---|---|
| Frontend UI / pages / components | mida-skills:mida-cms |
| API routes / services / models | mida-skills:mida-api |
| Heatmap / ClickHouse | mida-skills:mida-hm |
| Session recording | mida-skills:mida-recorder |
| Search / Elasticsearch | mida-skills:mida-search |
| MCP tools / AI assistant | mida-skills:mida-mcp |
| Shopify extension (liquid) | mida-skills:mida-extension |
| Proxy / auth routing | mida-skills:mida-proxy |
Most features touch both mida-cms (frontend) and mida-api (backend) — load both skills.
Combine all gathered context:
Then implement following the checklist in the relevant skill (e.g. mida-skills:mida-cms Add Feature Checklist).
| Step | What to tell the user |
|---|---|
| After Step 1 | Summary: task title, affected repos, Figma link found/not found |
| After Step 2 | UI spec: list of Polaris components to build, spacing summary |
| After Step 3 | Validation result: any component corrections needed |
| After Step 4 | Confirm patterns loaded, ready to implement |
| After Step 5 | List files created/modified, how to test |
If no Figma link in the Jira ticket:
/mcp → select claude.ai Atlassian Rovo → authenticatenpx claudepluginhub quyensatoru/mida-md --plugin mida-skillsProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.