docs(providers): align setup guidance

[deepujain]

Summary

• Align .env.example, the README provider reference, and the development setup guide with the provider selector on current main.
• Document the dedicated Ollama, Azure OpenAI, generic OpenAI-compatible, and Gemini CLI setup paths.
• Preserve existing provider behavior; this PR changes setup guidance only.

Validation

• Compared every documented provider name, credential variable, endpoint default, and model default against the corresponding implementation under src/skillspector/providers/.
• Verified all eleven usable provider names and the new provider-specific variables occur in .env.example, README.md, and docs/DEVELOPMENT.md.
• pytest -m 'not integration and not provider' tests/ -q — 2,825 passed, 13 skipped, 38 deselected, 4 xfailed.
• ruff check src tests — passed.
• ruff format --check src tests — 193 files already formatted.
• git diff --check — passed.
• Inspected the Markdown tables and fenced setup examples in source; no local Markdown renderer is configured in the repository.

Risk

• Low: documentation and the example environment template only; runtime provider selection and request behavior are unchanged.
• Live-provider tests were not run because no provider code changed.

Fixes #433

[rng1995]

[SkillSpector Review]

The added provider setup paths otherwise match their implementations, but the table would still advertise nv_build's retired default after merging into current main. Please update that model reference so the setup documentation is actually aligned. All required checks are green.

[rng1995]

[P2] Update the nv_build default while aligning this table. On current main, NvBuildProvider.DEFAULT_MODEL is z-ai/glm-5.2; the old deepseek-v4-flash is explicitly retired and returns 410. This PR merges cleanly without changing this row, so the newly aligned provider table would still advertise a dead default. Change this cell to z-ai/glm-5.2 and update any copied default references.