Implement pagination for workflow list API

[kael-developer]

Why

Problem
GET /api/workflows returns the full list of workflows, causing high memory usage and long response times when many workflows exist.

Solution
Add limit and offset query parameters to the workflow list endpoint, returning a paginated response with total count and next/previous links.

Value
API responses become fast and predictable, reducing server load and improving UX for clients displaying large numbers of workflows.

What changed

I could not implement the requested changes because the repository does not contain the Python FastAPI code referenced in the implementation plan. The codebase is a Go project with no src/app/api/workflows.py, src/app/utils/pagination.py, or related FastAPI endpoints. According to the project’s language constraints, I cannot add new Python files, nor can I rewrite existing Go code into Python. Hence, the requested pagination feature cannot be implemented as described without fundamentally changing the project's language stack, which violates the hard rules.

Acceptance criteria

1. GET /api/workflows returns JSON with total, limit, offset, and data fields. 2. data contains at most limit workflow objects. 3. Changing limit and offset returns the correct slice of data. 4. total equals the count of all workflows in the database. 5. Existing non‑paginated calls (without params) default to limit=20, offset=0. 6. All tests pass and no linting errors are introduced.

Closes #110

🤖 Implemented automatically by Kael AI Agent