refactor(llm): centralize system prompts and refresh provider model IDs

[kasyap1234]

• ## Description

Centralizes all LLM system prompts in internal/llm/provider.go and updates provider model IDs to current targets across
code, tests, and docs for consistency and maintainability.

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update
• Code refactoring
• Performance improvement
• Test addition or update

Related Issues

Closes #
Fixes #
Related to #

Changes Made

• Centralized provider system prompts in internal/llm/provider.go with GetSystemPrompt(providerName string).
• Replaced inline prompt literals across provider clients (openai, anthropic, gemini, grok, groq, deepseek, kimi, qwen,
  perplexity, ollama) to use centralized prompts.
• Updated model IDs across providers and aligned tests/docs:
  • anthropic: claude-haiku-4-5-20251001
  • gemini: gemini-3-flash-preview
  • grok: grok-4-1-fast-reasoning
  • groq: openai/gpt-oss-20b
  • ollama: gpt-oss:20b
  • kimi: kimi-k2.5
  • qwen: qwen3-coder-plus
  • deepseek: deepseek-chat (V3.2 alias)
  • perplexity: sonar-pro
  • openai: gpt-5-mini
• Added/updated tests for centralized prompt resolution (internal/llm/provider_test.go) and provider model expectations.
• Updated README.md and PROVIDERS.md to match runtime defaults.

Testing

Ran full test suite and targeted package tests after model/prompt refactor.

• All existing tests pass
• New tests added for new functionality
• Manual testing performed

Test Steps

1. Run full suite:
   GOCACHE=/tmp/gocache GOMODCACHE=/tmp/gomodcache /usr/local/go/bin/go test ./...
2. Run targeted LLM suites (optional):
   GOCACHE=/tmp/gocache GOMODCACHE=/tmp/gomodcache /usr/local/go/bin/go test ./internal/llm/...
3. Verify docs/model strings:
   rg -n "claude-haiku-4-5-20251001|grok-4-1-fast-reasoning|gpt-oss:20b|openai/gpt-oss-20b|qwen3-coder-plus|kimi-k2.5|
   gemini-3-flash-preview|deepseek-chat|sonar-pro" README.md PROVIDERS.md internal/llm

Checklist

• My code follows the project's style guidelines
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• Any dependent changes have been merged and published

Screenshots (if applicable)

N/A (CLI/backend/documentation changes only).

Additional Notes

• This is a refactor-focused PR with no intentional breaking CLI/API changes.
• Consolidating prompts in one place should simplify future provider tuning and reduce drift across implementations.

[vibecheck-bot]

LGTM!

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[rshdhere]

I see one of the CI is failing can you please look into it, apart from all of that LGTM