feat: Expose configurable version & health check API for Django integration

[ywatanabe1989]

Summary

Expose version checking and health status APIs from scitex package for reuse in Django (scitex-cloud) and other integrations. Single source of truth for ecosystem health monitoring.

Motivation

• scitex-cloud currently duplicates version checking logic
• Dashboard already has version checking in scitex._dev._versions
• Docker container health checks needed for dev/staging/prod monitoring
• Should be configurable for different deployment environments

Proposed API

1. Version Checking (already exists, needs public exposure)

from scitex.dev import get_ecosystem_versions

# Returns dict with installed versions for configured packages
versions = get_ecosystem_versions()
# {"scitex": "2.17.5", "figrecipe": "0.21.0", ...}

2. Docker Health Checker (new)

from scitex.dev import check_docker_health

# Check versions in Docker containers via HTTP endpoints
health = check_docker_health(endpoints={
    "dev": "http://127.0.0.1:8000/api/versions/",
    "staging": "http://staging.scitex.ai/api/versions/",
    "prod": "https://scitex.ai/api/versions/",
})

3. Configuration File Support

# ~/.scitex/health.yaml or pyproject.toml [tool.scitex.health]
ecosystem_packages:
  - scitex
  - figrecipe
  - crossref-local
  - openalex-local
  - scitex-writer
  - socialia

docker_endpoints:
  dev: http://127.0.0.1:8000/api/versions/
  staging: http://staging.scitex.ai/api/versions/
  prod: https://scitex.ai/api/versions/

Dashboard Integration

• Add DOCKER columns: dev | staging | prod
• Fetch from configured endpoints
• Show version alignment across all environments

Priority

Medium - Useful for deployment monitoring but not blocking

[ywatanabe1989]

Update: Hub-and-Spoke Architecture Decision

After further discussion, we've decided on a consistent ecosystem architecture:

Pattern: scitex as Hub, Delegating to Spoke Packages

                    ┌─────────────────────┐
                    │   scitex (hub)      │
                    │   import scitex as stx
                    └─────────┬───────────┘
                              │
        ┌─────────┬───────────┼───────────┬─────────┐
        ▼         ▼           ▼           ▼         ▼
   stx.cloud  stx.writer  stx.scholar  stx.plt  stx.stats
        │         │           │           │         │
        ▼         ▼           ▼           │         │
  ┌─────────┐ ┌─────────┐ ┌─────────┐    │  (core) │
  │scitex-  │ │scitex-  │ │crossref-│    │         │
  │cloud    │ │writer   │ │local    │    │         │
  │(Django) │ │         │ │openalex-│    │         │
  └─────────┘ └─────────┘ │local    │    └─────────┘
                          └─────────┘

Implementation

scitex/cloud/init.py (thin delegation):

try:
    from scitex_cloud import *  # delegate to actual package
except ImportError:
    raise ImportError("Install scitex-cloud: pip install scitex-cloud")

User Experience

import scitex as stx

stx.cloud.get_version()    # → scitex-cloud package
stx.cloud.health_check()   # → scitex-cloud package
stx.writer.compile()       # → scitex-writer package
stx.scholar.search()       # → crossref-local/openalex-local

Benefits

• Consistent with existing ecosystem (scitex-writer, figrecipe, etc.)
• Each spoke package versioned independently
• Main scitex package stays lightweight
• Optional dependencies - graceful failure if spoke not installed

Next Steps

1. Create scitex/cloud/ delegation module in scitex-python
2. scitex-cloud package exports version/health APIs via scitex_cloud
3. Version dashboard fetches via stx.cloud or direct HTTP endpoints