fix(service-offerings): qualify Enterprise local inference claims

- Change 'no API dependency' to 'API fallback available for large models'
- Update package tagline to 'local inference capability'

Closes #907

operation-get-a-job/service-offerings.md

@@ -138,10 +138,10 @@ Whitestone Engineering is a human-led, AI-augmented engineering firm. Our princi
 ---
 
 ### Enterprise — $40,000+
-*Full sovereign infrastructure with local inference.*
+*Full sovereign infrastructure with local inference capability.*
 
 - Full agent fleet (5+ instances)
-- Local LLM inference stack (no API dependency)
+- Local LLM inference stack (API fallback available for large models)
 - Self-hosted code forge (Gitea) with CI/CD
 - Agent security hardening and conscience validation
 - Nostr-based sovereign communication layer