Awesome Ai Agents 2026
Section: Context Optimization · Context engineering engine. 100% codebase visibility with 78% fewer tokens. Knapsack-optimal selection, SimHash dedup, RL from response quality. Rust engine, <10ms. MCP + HTTP proxy.
Entry
Appears in 5 awesome lists
Context engineering engine. 100% codebase visibility with 78% fewer tokens. Knapsack-optimal selection, SimHash dedup, RL from response quality. Rust engine, <10ms. MCP + HTTP proxy.
Section: Context Optimization · Context engineering engine. 100% codebase visibility with 78% fewer tokens. Knapsack-optimal selection, SimHash dedup, RL from response quality. Rust engine, <10ms. MCP + HTTP proxy.
Section: MCP Servers and Directories · Local-first MCP server for explicit-budget context selection, content-addressed exact recovery, and auditable Context Receipts. Apache-2.0.
Section: Optimizations · Information-theoretic context optimization proxy. Cuts LLM token costs by 70–95% with zero accuracy loss using greedy submodular knapsack maximization.
Section: 13. Developer Tools & Integrations · Local-first MCP server for explicit-budget context selection, content-addressed exact recovery, and auditable Context Receipts. Apache 2.0 licensed.
Section: Artificial Intelligence · Information-theoretic Context Engineering Engine that uses reinforcement learning to intelligently prune and select optimal RAG fragments.
High-performance neural network inference framework optimized for mobile platforms. No third-party dependencies, cross-platform, and runs faster than all known open-source frameworks on mobile CPU. Powers Tencent apps including QQ, WeChat, and Pitu. BSD-3-Clause licensed.
Supercharge LLM inference with the fastest KV Cache layer. 3-10x delay savings and GPU cycle reduction for multi-round QA and RAG. Integrates seamlessly with vLLM for distributed, high-throughput deployments. Apache 2.0 licensed.
LangWatch is a visual interface for DSPy and a complete LLM Ops platform for monitoring, experimenting, measuring and improving LLM pipelines, with a fair-code distribution model.
Long-term memory for Gemini CLI and other MCP clients over a local stdio MCP server. Five tools (remember, recall, list_memories, forget, fit_context): messages are stored word for word in local SQLite or your own Postgres, and recall returns the stored messages that answer a question within a…
Cognitive-harness MCP server with four tools (reasoning, code, anti-deception, memory) returning a structured scaffold (failure pattern, procedure, suppression vectors, falsification test) the agent absorbs before generating. Hosted at api.ejentum.com/mcp and on npm as ejentum-mcp.
The Context OS for AI Development. Reduce token waste in Cursor, Claude Code, Copilot, Windsurf, Codex, Gemini & more by 60–95% (up to 99% on cached reads) Shell Hook + MCP Server · 49 tools · 10 read modes · 90+ patterns · Single Rust binary
MCP server that generates production-ready visual assets (app icons, favicons, OG images, logos, wordmarks) by routing requests across 30+ image generation models. Zero API key required for first run via Pollinations and Stable Horde free tiers. Works with any MCP client including Gemini CLI.
Remote MCP server for AI coding assistants with real-time news and multi-dimensional bias scoring (3.2M+ articles, 5000+ sources), financial markets (stocks, ETFs, crypto, AI bull/bear cases), ML options pricing (ITM probability, Greeks, fair value), balanced news synthesis, trending topics, and…