feat: enhance silva AI compatibility

[craigyu]

Summary

This PR adds structured repository and frontend guidance for AI-assisted development, improving consistency and reducing incorrect assumptions across the Silva project.

It introduces repo-level instructions for how agents should work within the codebase, reinforces project conventions, and adds frontend-specific guidance for React, TypeScript, styling, and testing patterns. The goal is to make multi-agent and AI-assisted workflows more predictable and easier to maintain.

Changes

• Added a top-level project guide in AGENTS.md covering:

  • project overview and domain context
  • backend and frontend responsibilities
  • full-stack development workflow
  • testing and quality expectations
  • guidance for clarifying domain assumptions before implementation

• Added Copilot instruction metadata in .github/copilot-instructions.md to enable the repository’s AI workflow and skill setup.

• Added frontend contributor guidance in frontend/AGENTS.md, including:

  • screen and component conventions
  • naming and folder structure patterns
  • styling and import conventions
  • implementation and testing expectations for the React app

• Updated the Caveman skill documentation to align with the repo-specific instruction setup.

Impact

This is a documentation and workflow enhancement only. There are no changes to runtime application behavior or business logic.

---

Thanks for the PR!

Deployments, as required, will be available below:

• Hybrid Backend
• Hybrid Frontend
• Postgres Backend
• Postgres Frontend

Please create PRs in draft mode. Mark as ready to enable:

• Analysis Workflow

After merge, new images are deployed in:

• Merge Workflow

[Copilot]

Pull request overview

Adds repo-level and frontend-specific agent guidance to make AI-assisted development more consistent across Silva (domain context, workflows, conventions, testing expectations), plus Copilot instruction metadata and skill documentation alignment.

Changes:

• Added top-level AGENTS.md documenting Silva domain context, stack, and full-stack workflow expectations.
• Added frontend/AGENTS.md documenting React/TypeScript/Carbon conventions, styling rules, API usage patterns, and testing structure.
• Added Copilot instruction entry point and updated Caveman/Ponytail skill docs to align with the repo’s agent setup.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
frontend/AGENTS.md	New frontend agent conventions for screens/components, styling, API usage, and tests
AGENTS.md	New top-level Silva overview + cross-layer workflow/testing expectations
.github/copilot-instructions.md	Enables Copilot instruction metadata pointing to skill docs
.agents/skills/ponytail/SKILL.md	Adds Ponytail skill doc (YAGNI/lazy-dev ladder)
.agents/skills/caveman/SKILL.md	Updates Caveman skill doc metadata and wording

Suppressed comments (4)

frontend/AGENTS.md:768

• This section says the entire src/services/ folder is auto-generated and that only API.ts should be modified. In this repo, only src/services/OpenApi/** is generated; API.ts, AuthService.ts, and OpeningSearchService.ts are hand-written. Clarifying prevents accidental edits to the wrong files.

The `/src/services/` folder is **auto-generated** from the backend's OpenAPI specification.

### Rules

- **Never modify auto-generated files** (e.g., individual service files)

frontend/AGENTS.md:843

• Unit testing guidance says “Use Jest”, but the repo is set up to run unit tests with Vitest (and uses @testing-library/jest-dom matchers via src/setupTests.ts). Also, the example path uses __tests__ while this repo uses src/__test__/.

Use **Jest** and **React Testing Library** for component tests:

```typescript
// src/components/Avatar/__tests__/Avatar.test.tsx
import { render, screen } from '@testing-library/react';

frontend/AGENTS.md:718

• The remainder of this example continues with API.updateOpening(...) / cache updates for an openings list, but those methods aren’t available on the API export here. Rewriting the rest of the snippet to match a real endpoint keeps the guidance copy/paste-safe.

  const updateMutation = useMutation({
    mutationFn: (opening: Opening) =>
      API.updateOpening(opening.id, opening),
    onSuccess: (updatedOpening) => {
      // Invalidate cache

frontend/AGENTS.md:478

• This calls out “Jest configurations” as the reason @/ might not resolve in tests, but the frontend test runner here is Vitest and there’s no Jest config in the repo. Suggest generalizing this to “test runner configuration” (or Vitest specifically) to match reality.

- **Exception in tests:** Relative paths may be necessary if `@/` doesn't resolve (e.g., in Jest configurations)

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[Abradat]

🤖🚀

[paulushcgcj]

The PR does a good job of keeping the Copilot entry point small and putting the detailed behavior in focused skills. That separation reduces duplication and gives us a cleaner place to evolve the instructions over time.

[craigyu]

Paulo's comment are fixed, will merge this in for now