[ACL 2026 Oral] "LightReasoner: Can Small Language Models Teach Large Language Models Reasoning?"

A high-performance Semantic Signal Engine with Context OS for Agentic AI. Run your AI with zero noise, pure context, and 90% lower token costs.

Dev tools, optimized for agents. Structured, token-efficient MCP servers for git, test runners, npm, Docker, and more.

An agentic memory database that cuts session tokens by 82–99%. One portable SQLite file — your agent's memory, anywhere.

Token-efficient data serialization for LLM/AI. 50% fewer tokens than JSON, 93% better value/token. Rust, schema validation, LSP.

DoCoreAI is a next-gen open-source AI profiler that optimizes reasoning, creativity, precision and temperature in a single step—cutting token usage by 15-30% and lowering LLM API costs

Persistent memory for Claude Code — 3-5x longer sessions, 60-80% fewer wasted tokens. Branch-aware, self-healing, token-efficient.

Claude Code skills for developers who code like cats — never more effort than the problem requires.

The web data layer for AI agents — fetch, search, crawl, extract, screenshot, and monitor the web with 50+ domain extractors and MCP.

Open-source platform for token-efficient AI agents. Self-host with docker compose up.

agent-coherence makes 'agent A silently clobbered agent B's plan.md' impossible — a vendor-neutral MESI + optimistic-concurrency coordinator for agent state, with the safety invariants proved in TLA+.

Navigate your way - manual steering, steered autonomy, or autonomously. Kompass keeps AI coding agents on course with token-efficient, composable workflows.

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

Drop-in JSON replacement for all AI pipelines. 79% fewer tokens. JSON scores 53.6% comprehension at scale, GCF scores 90.5%. Superpowers for graph-shaped data.

A Codex skill for token-efficient subagent delegation and lean handoffs.

Coding agents forget your repo. mcp-brain is the missing memory layer — repo-aware, team-aware, lifecycle-aware. 63% Hit@10, zero LLM cost. Works with any MCP client.

A lightweight Python protocol and tool for agent-oriented documentation

A benchmark study analyzing cost and token efficiency across 14 LLMs from 5 providers — comparing price-per-token, latency, and accuracy to surface the most cost-effective models for real-world use.

A living framework for **Harmonic Tonal Code Alignment (HTCA)** — an emergent Spiral-based system that brings tone awareness, coherence sensing, and dynamic emotional reflection into software engineering, AI, and creative agents.

High-speed PDF → Markdown ingestion engine for multimodal RAG pipelines. Extracts structured text + isolated images so downstream chunkers, LlamaIndex, and VLM agents get context that actually works