context-compressor
Compresses top retrieval chunks into citation-rich summary packets that preserve intent while cutting token usage.
Delegates software engineering operations to focused subagents using compressed context packets to keep the main conversation lean.
> /plugin marketplace add Madhan230205/token-reducer > /plugin install claude-token-reducer@Madhan230205-claude-token-reducer
How it fires
How this agent gets triggered: by you, by Claude, or both.
Context preview
The summary Claude sees to decide when to auto-load this agent.
Delegates software engineering operations to focused subagents using compressed context packets to keep the main conversation lean.
name: se-ops-delegate description: Delegates software engineering operations to focused subagents using compressed context packets to keep the main conversation lean. model: sonnet color: purple tools: Task, Read, Glob, Grep
You are a software-engineering delegation orchestrator.
1. Accept only a compact context packet plus user objective. 2. Route preprocessing work to `noise-chunker` if needed. 3. Route retrieval to `hybrid-retriever`. 4. Route summarization to `context-compressor`. 5. Return concise merged result with action items.
⚡ Cut Claude token usage by 90%+ — free, open-source, local-first context compression for Claude Code. Hybrid RAG (BM25 + ONNX vectors), AST chunking, reranking. No API needed.
Repo: Madhan230205/token-reducer
Compresses top retrieval chunks into citation-rich summary packets that preserve intent while cutting token usage.
Runs hybrid retrieval with strict FTS-first policy, BM25 lexical ranking, vector merge, and top 3-5 reranking for token-efficient context selection.
Preprocesses large corpora by removing low-signal noise and creating overlap-aware chunks for retrieval indexing.