MCP servers and skills: search, filter by where they run, and compare.
1–24 of 604 tools for “long-context”
Clear allLong-context MoE training guidance for Megatron Bridge. Covers CP sizing, selective recompute, dispatcher choices, and practical patterns from DSV3, Qwen3, and Qwen3-Next long-context experiments.
OfficialSkillApache-2.0★ 3.6k in collectionNo key
Validate and use packed sequences and long-context training in Megatron-Bridge, including offline LLM packing, collate-time VLM packing, Energon online packing, and CP constraints.
OfficialSkillApache-2.0★ 3.6k in collectionNo key
Cognee is the open-source AI memory platform for agents. Give your AI agents persistent long-term memory with small models for free
in the skill textservice. ## Benchmarks and research The BEAM evaluation measures conversational memory using synthetic long-context conversations and an LLM judge. The reported runs use Cognee's memory
SkillApache-2.0★ 32kNo key
Agentic-first Cursor Rules powered by MiniMax M3 - clarify-first prompting, interleaved thinking, and full tool orchestration for production-ready AI coding
in the readmethe execution spine — reasoning protocol, solver loop, scope control, code discipline, M3 long-context discipline, M3 multimodal input discipline, and a strict proof contract. No persona
SkillMIT★ 126No key
Messy long context -> structured insight -> auditable markdown, research threads, and inspectable outcomes.
PythonMIT★ 1Few starsNo key
Hivemind turns your traces into reusable skills across agents
in the skill textagent on the team can execute the pattern. On LoCoMo , the public long-context memory benchmark, Hivemind is **25% cheaper, 1.7× fewer tokens, and 31% fewer
SkillApache-2.0★ 1.6kNo key
Atomic Agent is a local-first AI agent. Runs open-weight models on your own machine via llama.cpp.
in the readmeversus F16, with a fused Metal decode kernel, so long-context sessions fit in far less memory. - **TurboQuant weights:** Lloyd-Max weight quantization with WHT rotation
Node.jsMIT★ 3.4kNo key
Claude Code skill for handling long-context tasks through recursive decomposition
SkillMIT★ 52No key
Which 5-hour window took you out — and where the money goes. Local profiler for Claude Code, OpenClaw & Hermes Agent: usage windows, burn by source (cron/chats/subagents), the overnight bill, why it b
in the readmeusage window, and how long until the wall uvx agentburn context # what long contexts cost — and what a /clear at 150k would have saved ``` ## Two ways
PythonMIT★ 113No key
Whale — blazingly fast, terminal-first AI coding agent for DeepSeek. ~98% prompt cache hit rate, 1M context, MCP tools, dynamic workflows.
in the readmehit · Zero bloat Whale — AI coding agent for DeepSeek, in any environment. Long context, tools, and programmable workflows — start in the terminal, scale to desktop and
GoMIT★ 933No key
Crow — a local AI agent, not a chat box: 30 tools, MCP, memory, skills, goals, subagents, vision, and image generation and editing with Qwen-Image. Runs on crow-nest or llama.cpp on one GPU, or any mo
PythonMIT★ 37No key
Re-anchor long conversations that drift through repeated local corrections.
SkillMIT★ 102No key
Self-hosted AI agent memory — MCP memory server with durable recall for agents
in the readmeSimilarity search | Ingestion, MCP tools, durable agent memory, sparse + graph retrieval | | Long context | More tokens in one prompt | Persistent memory across sessions, scoping, freshness | | Chat history
PythonApache-2.0★ 107No key
面向 Codex / Claude Code 的中文长篇网络小说写作 Skill：分层上下文组装、50章分阶段规划、状态回证、伏笔追踪与回收、一致性审查和自动修复。Chinese web novel writing agent skill.
SkillMIT★ 70No key
TOMC Memory: CPU context compression and task memory for Codex, Claude Code, Cursor, TRAE, WorkBuddy and other MCP clients.
in the readmesupport the project. TOMC Memory Context compression and task memory for Codex / Claude Code / Cursor / TRAE / WorkBuddy Compress and compile long histories, and recall saved task
PythonMIT★ 15No key
A Model Context Protocol (MCP) server that provides advanced code analysis and reasoning capabilities powered by Google's Gemini AI
in the readmebottlenecks - **Hypothesis Testing**: Tests theories about code behavior with evidence-based validation - **Long Context Support**: Leverages Gemini 2.5 Pro Preview's 1M token context for analyzing
SkillMIT★ 109No key
An Agent Skill: drive Cline CLI coding tasks as the user's proxy.
in the skill textreal multi-module production backend): 1. **Initialization — use a strong long-context (paid) model.** Give it the full weight: whole-project codebase scan, writing project rules
SkillMIT★ 50No key
Read-only MCP server for official Python docs: local index, no API keys, exact symbol lookup, version-aware retrieval.
in the readmeyou actually ship on. ## Why this exists There is a difference between long context and usable context. Research like the *Lost in the Middle* study (Liu
PythonMIT★ 9No key
Shared project knowledge and handoff for coding agents across long-running projects
in the readmeone MCP. **Chat windows were for thinking.** ChatGPT and Claude could keep long-context brainstorming, product decisions, and difficult design conversations in the conversational surface where
PythonMIT★ 17No key
MCP Deep Research Server using Gemini creating a Research AI Agent
in the readmeContext Protocol (MCP) server/tool for seamless agent integration. * **Gemini 2.5 Flash Pipeline:** Long-context reasoning, structured JSON outputs, and tool use (Google Search Grounding, Code Execution
SkillMIT★ 69No key
Engineering best practices that AI coding agents actually apply — 40+ skills, from secure coding and cloud to compliance and UX writing. BUILD + AUDIT modes. Measured: best-practice coverage 59% → 98%
in the skill textand re-check them *last* before shipping — so the guidance survives a long context instead of fading into it. That's why it beats a bigger
SkillCC-BY-4.0★ 21No key
ProbeCode: AI coding agent integrating static code inspection with a ReAct framework to understand and memorize long-context code.
PythonMIT★ 4Few starsNo key
MCP server giving Claude Code long-context memory of a codebase through Gemini's 1M-token window: one workspace scan, implicit caching by default, explicit Context Cache on request (125 s → 14 s measu
Node.jsMITFew starsNo key
A high-performance, local, offline-first AI memory system built in Rust
in the readmeStandards** for AI memory. Our methodology prioritizes high-integrity reasoning and ultra-long context persistence over synthetic "recall-only" metrics. ### 2026 Gold Standard Validation _Verified multi
SkillMIT★ 37No key
Who stands behind it
