prompt-compression

14-stage Fusion Pipeline for LLM token compression — reversible compression, AST-aware code analysis, intelligent content routing. Zero LLM inference cost. MIT licensed.

Headroom for macOS — cut Claude Code and Codex token costs by ~50%

Drop-in prompt compression for production LLM apps. Cut your token bill 40-60% without changing your code. Python SDK, LLMLingua-2, MIT.

Local proxy that compresses your LLM API requests so you pay less, with no change to the answers. Trims wasted tokens from prompts, history, tool output, and code before they're sent: -31% input / -74% output, measured live. Any provider, no extra model calls. Also an MCP server and embeddable library (Rust, Python, Ruby, Kotlin, Swift, JS/TS).

HEWN 2.0 2026: AI Output Router for Precision Summaries & Polished Code

CPU-first LLM inference server on llama.cpp. Runs useful models on free-tier ARM boxes; rewriting the input made it ~6x faster and more accurate than tuning the engine. MIT, benchmarks and failures included.

A curated list of strategies, tools, papers, and resources for reducing LLM token costs and improving efficiency in production.

Claude Code & Cursor token usage monitor and cost tracker for macOS & Windows. Live token/cost/burn monitoring across 8 AI coding agents, a budget circuit breaker that stops runaway agents before the next API call, an MCP manager, and 40-70% on-device prompt compression.

JavaScript/TypeScript implementation of LLMLingua-2 (Experimental)

Rolling context compression for Claude Code — never hit the context wall. Auto-compresses old messages while keeping recent context verbatim. Zero config, zero latency. Works as a Claude Code plugin.

A self-improving knowledge base about LLM agent infrastructure

VL-JEPA inspired pipeline — compress images/text locally via Ollama, send compact payloads to any LLM API. Cut token costs by ~80%.

A lossless token-optimization codec for LLM agents — cut input tokens up to 82%, nothing dropped, every fact kept. Drop-in proxy + one-line Python API for Claude Code, Codex, Cursor, OpenAI, Anthropic & Ollama. pip install foveance

Python command-line tool for interacting with AI models through the OpenRouter API/Cloudflare AI Gateway, or local self-hosted Ollama. Optionally support Microsoft LLMLingua prompt token compression

Lossless-first prompt compression for JSON, YAML, CSV, and Markdown. Library, CLI, MCP server, desktop app, and browser extension.

Reverse T9 for LLMs. Free, open-source prompt compressor for your AI prompts and agents.

Caveman prompting, measured. A two-channel evaluation protocol scoring what input and output compression actually cost LLMs in dollars, accuracy, and surface-text fidelity across seven models and five benchmarks.

GenPark AI Agent Skill - Multi-tier model cascading router dynamically selecting optimal models (e.g. Small vs Medium vs Large) based on task complexity, cost budgets, and SLA constraints.

GenPark AI Agent Skill - Mixture-of-Agents (MoA) multi-layer aggregator synthesizing diverse candidate proposals from heterogeneous sub-agents into high-consensus outputs.

GenPark AI Agent Skill - Radix tree prefix caching simulator for prompt templates and agent system instructions, maximizing KV-cache hit rates and cutting TTFT latency.