mirror of
https://github.com/headroomlabs-ai/headroom.git
synced 2026-08-27 14:17:10 -04:00
3ebcd89d46
README: 694 → 203 lines. Crisp, scannable, links to docs. - compress() as the hero quickstart (not proxy) - Integration table: compress(), LiteLLM, ASGI, proxy, Agno, LangChain - LangChain marked as experimental - "Already have a proxy?" callout linking to Integration Guide - Architecture: ContentRouter (not SmartCrusher) as the primary compressor New: docs/integration-guide.md - Detailed setup for every integration path - compress() with Anthropic, OpenAI, LiteLLM, raw HTTP - LiteLLM callback + LiteLLM proxy ASGI middleware - ASGI middleware for any FastAPI/Starlette app - Compression hooks for advanced customization - FAQ section Fix: compress() uses default pipeline (CacheAligner + ContentRouter + IntelligentContext) instead of manually specifying SmartCrusher. |
||
|---|---|---|
| .. | ||
| plans | ||
| agno.md | ||
| api.md | ||
| ARCHITECTURE.md | ||
| benchmarks.md | ||
| ccr.md | ||
| compression.md | ||
| configuration.md | ||
| errors.md | ||
| getting-started.md | ||
| image-compression.md | ||
| index.md | ||
| integration-guide.md | ||
| langchain.md | ||
| llmlingua.md | ||
| macos-deployment.md | ||
| mcp.md | ||
| memory.md | ||
| metrics.md | ||
| proxy.md | ||
| quickstart.md | ||
| sdk.md | ||
| text-compression.md | ||
| transforms.md | ||
| troubleshooting.md | ||