2 Commits

Author	SHA1	Message	Date
oyi77	f99c90dc85	feat(cache): add OpenAI prompt_cache_key and Gemini cachedContent support ... Provider-specific caching enhancements: OpenAI: - Auto-generate prompt_cache_key from message prefix hash - Key format: omni-{prefix_hash_32chars} - Preserves client-provided keys (doesn't override) Gemini: - Preserve cachedContent ID if provided by client - Enables explicit Gemini caching for long prompts - cachedContentTokenCount already tracked in responses	2026-03-31 04:50:50 +07:00
oyi77	6780485051	feat(cache): persistent metrics, cache entry browser, settings UI, MCP tools, prefix analyzer ... Implements remaining features from #813: Phase 1 - Persistent Metrics: - Add cache_metrics table for persistent hit/miss tracking - Semantic cache stats now survive server restarts Phase 2 - Cache Entry Browser: - /api/cache/entries endpoint with search, pagination, delete - CacheEntriesTab component for browsing cached entries Phase 3 - Settings UI: - CacheSettingsTab for semantic/prompt cache configuration - /api/settings/cache-config endpoint Phase 4 - Prefix Analyzer: - src/lib/promptCache/prefixAnalyzer.ts for intelligent caching - Analyzes message arrays to find stable prefixes Phase 5 - Provider Support: - Added deepseek to CACHING_PROVIDERS Phase 6 - MCP Tools: - omniroute_cache_stats tool - omniroute_cache_flush tool Phase 7 - Retention: - cleanOldMetrics() for auto-purge of old entries Closes #813	2026-03-30 17:32:45 -03:00