Stop building your AI stack from 40 open tabs.

DeusData 34.2k · yesterday

Code intelligence MCP in pure C: tree-sitter knowledge graph over 158 languages, average repo indexed in milliseconds, sub-ms queries, 10x fewer tokens. Single static binary, zero deps.

phuryn 24.3k · 20 days ago

PM Skills Marketplace: 68 skills and 42 chained workflows in 9 plugins — discovery, strategy, PRDs, launch, growth — encoding Torres/Cagan-style frameworks for Claude Code and Cowork.

DietrichGebert 87.9k · 8 days ago

A skill that makes your agent code like the laziest senior dev: YAGNI enforced — ~54% less code, ~20% cheaper, ~27% faster on measured Claude Code sessions. Works with 20 agents.

zilliztech 2.3k · yesterday

Zilliz's unified memory for coding agents: one Markdown + Milvus store shared across Claude Code, Codex, OpenCode and OpenClaw — hybrid search, plus repeated workflows distilled into skills.

hexo-ai 2.1k · 21 days ago

Self-improving loop from the SIA paper: Meta, Target and Feedback agents evolve a task agent's harness AND weights against a benchmark — #1 on MLE-Bench Hard, 14x kernel speedups.

FareedKhan-dev 8.5k · 29 days ago

The full LLM pipeline hand-written in plain PyTorch — tokens, transformer, pretraining, then SFT, reward model, PPO, DPO, GRPO. No trl, no peft: read every algorithm, train on one GPU.

cocoindex-io 2.6k · 5 days ago

AST-based semantic code search for coding agents: pipx install, zero config, local embeddings out of the box — a CLI/skill/MCP that cuts agent context ~70% vs grepping. Built on CocoIndex.

alibaba 11k · today

Alibaba's battle-tested AI code-review CLI: deterministic pipelines + LLM agent, line-level comments, tuned rulesets (NPE, XSS, SQLi) — higher precision than general agents at ~1/9 the tokens.

kayba-ai 2.5k · 15 days ago

Learning loop for any agent: reflect on failures, distill strategies into a Skillbook, inject them next run — 2x consistency on Tau2, 49% token cuts. LiteLLM-based, 100+ providers.

confident-ai 17.1k · yesterday

Pytest for LLM apps: 40+ research-backed metrics — G-Eval, RAG suite, agent task completion, hallucination — as unit tests you run in CI, judged by any LLM including local ones.

headroomlabs-ai 61.3k · today

Context compression layer for agents: squeezes tool outputs, logs, files and RAG chunks 20-95% before the LLM — reversible, local-first; library, proxy, one-command agent wrap, or MCP.

lfnovo 35.9k · yesterday

Self-hosted NotebookLM alternative: multi-modal sources, vector plus full-text search, context-aware chat and multi-speaker podcast generation — 18+ model providers incl. Ollama, full REST API.

evo-hq 1.3k · 6 days ago

Karpathy-style autoresearch on any codebase: /evo:discover instruments the benchmark, /evo:optimize runs tree search with parallel subagents in worktrees. Plugin for Claude Code, Codex & co.

Goekdeniz-Guelmez 250 · 7 days ago

Native Mac app for on-device LLM fine-tuning via mlx-lm-lora: pick a model, choose SFT/LoRA/DPO-family algorithms, watch loss fall live, push to Hugging Face. No cloud, no code.

kennss 785 · yesterday

Sudoless Apple Silicon monitor: SwiftUI dashboard plus menu-bar suite tracking ANE, Media Engine and memory bandwidth Activity Monitor won't show — with DVR-style record & replay.

vectorize-io 18.7k · today

Agent memory that learns, not just recalls: retain/recall/reflect API over Postgres, SOTA on LongMemEval. Self-host via Docker with UI; Python/TS clients, any LLM provider.

StarTrail-org 7.1k · 7 days ago

Berkeley's visual RAG: render pages and PDFs to screenshot tiles and retrieve with a VLM embedder — tables, charts and layout survive. pixelshot CLI plus a hosted 8.28M-page Wikipedia index.

cjpais 27.2k · yesterday

Push-to-talk offline dictation: hotkey, speak, text lands in whatever field has focus. Whisper or Parakeet fully on-device; cross-platform Rust/Tauri, built to be forked.

omnigent-ai 7.7k · today

Open-source meta-harness over Claude Code, Codex, Cursor, OpenCode, Hermes, Pi and custom agents: swap harnesses without rewriting, enforce policies and sandboxes, follow sessions from any device.

LMCache 10.8k · today

KV-cache layer for scalable LLM serving: offload and reuse KV across GPU/CPU/disk/remote tiers to cut TTFT and prefill cost. vLLM-first; used by NVIDIA Dynamo and llm-d.

braedonsaunders 4.7k · 9 days ago

Paste a GitHub URL or drop a local folder → interactive architecture map in the browser: dependency graph, blast radius, health grade, security scan. Single index.html, zero install.

JustVugg 18.1k · today

Pure-C, zero-dep MoE runtime that runs GLM-5.2 (744B) on a 25GB-RAM consumer box by streaming experts from disk — VRAM/RAM/NVMe as one tiered hierarchy, never touching precision.

jbarrow 1.2k · 1 months ago

Turns any PDF into a fillable form: FFDNet models detect text, checkbox and signature fields; one CLI command writes the interactive PDF. Paper, dataset and weights all open.

katanemo 6.9k · yesterday

AI-native Envoy-based proxy for agentic apps: agent orchestration via a 4B routing model, smart LLM routing, guardrail filter chains and zero-code OTEL traces. Rust, framework-agnostic.

EXXETA 225 · 3 days ago

Local-first agentic runtime with persistent AI rooms and approval-gated memory: every memory write needs your OK; rooms, KB and artifacts are plain files on disk.

Canner 16.6k · today

Open-source GenBI engine: agents write governed SQL and deploy shareable dashboards over 22+ data sources, grounded in a Git-friendly context layer (MDL semantics, definitions, memory).

Mesh-LLM 2.8k · today

Distributed LLM inference in Rust: pool GPUs across machines into one OpenAI-compatible endpoint — local fit first, mesh routing, and stage splits for models too large for any single box.

BerriAI 54.4k · today

Open-source AI gateway: call 100+ LLM providers in OpenAI format via a Python SDK or self-hosted proxy — with cost tracking, virtual keys, guardrails, load balancing and logging.

zjunlp 1k · 7 days ago

ICLR 2026 memory framework for LLMs/agents: LLMLingua pre-compression, topic segmentation and offline memory updates — leading LoCoMo/LongMemEval results at lower token cost.

anthropics 589 · 2 months ago

Anthropic's harness primitives for long-running Claude agents: default-FAIL evidence gates, a fresh-context evaluator subagent and handoff hooks — each one standalone, readable file.

tirth8205 25.6k · 2 days ago

Local-first code intelligence graph for AI coding tools: Tree-sitter AST graph + blast-radius analysis served over MCP, so agents read ~82x fewer tokens per review question.

patchy631 36.7k · 8 days ago

36k-star hub of runnable AI engineering tutorials — LLMs, RAG and agent apps as self-contained projects, including build-code-harness: a Claude-Code-style coding harness rebuilt on CrewAI + E2B.

agno-agi 642 · 13 days ago

Company intelligence agent that navigates Slack, Drive, wiki and CRM live — no ingest/embed pipeline — and builds its own wiki + CRM as it learns your company.

danielmiessler 16.9k · yesterday

Daniel Miessler's LifeOS: an AI 'life operating system' that carries your goals and context into every task — an intent engineering platform with dashboard, agents and installer.

rituraj-io 4 · 3 months ago

Structured long-term memory over MCP: an enforced bucket-thread-memo hierarchy in one SQLite file, hybrid FTS5 + vector search fused with RRF, local embeddings.

InsForge 12.4k · today

All-in-one open-source backend for agentic coding: Postgres, auth, storage, edge functions, model gateway and site hosting — your coding agent operates it over MCP.