{"generated_at":"2026-10-21T05:23:46.782Z","refs":[
{"id":"gh:1jehuang/agentgrep","kind":"repo","title":"agentgrep","why":"Code search CLI built for agents, with a query DSL and trace-mode region expansion.","keys":{"races":["agents/code-intel"],"capabilities":["agents/code-intel/code-search"]},"lang":"Rust","license":"none","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:1jehuang/jcode","kind":"repo","title":"jcode","why":"Rust coding-agent harness built for heavy multi-agent workflows, with a TUI and MCP.","keys":{"races":["agents/coding"],"capabilities":["agents/coding/cli-agents"]},"lang":"Rust","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:1jehuang/jcode-desktop","kind":"repo","title":"jcode-desktop","why":"Native spatial desktop client for jcode, with voice input.","keys":{"races":["agents/clients"],"capabilities":["agents/clients/desktop-apps"]},"lang":"Rust","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:abdulmominsakib/localmind","kind":"repo","title":"localmind","why":"Flutter chat app for on-device models, local LLM servers and cloud providers.","keys":{"races":["llm-infra/chat-apps"],"capabilities":["llm-infra/chat-apps/mobile-chat"]},"lang":"Dart","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:abdulmominsakib/neural_tts","kind":"repo","title":"neural_tts","why":"On-device neural TTS for Flutter (Kokoro, Kitten, Supertonic) through native ONNX Runtime.","keys":{"races":["media/speech"],"capabilities":["media/speech/text-to-speech"]},"lang":"Dart","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:abiosoft/colima","kind":"repo","title":"colima","why":"Docker, containerd and k3s on macOS through Lima VMs, with minimal setup.","keys":{"races":["infra/containers"],"capabilities":["infra/containers/macos-containers"]},"lang":"Go","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:addyosmani/agent-skills","kind":"repo","title":"agent-skills","why":"Production-grade engineering skills and personas for coding agents.","keys":{"races":["agents/skills"],"capabilities":["agents/skills/engineering-skills"]},"lang":"JavaScript","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:agent-substrate/substrate","kind":"repo","title":"substrate","why":"Secure-by-default runtime that multiplexes millions of agent sandboxes onto Kubernetes pods.","keys":{"races":["agents/sandboxes"],"capabilities":["agents/sandboxes/sandbox-fleets"]},"lang":"Go","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:ai-sorcery/ai-sorcery","kind":"repo","title":"ai-sorcery","why":"Claude Code toolkit: install and update, commit guards, test-fixture capture and LLM-call routing.","keys":{"races":["agents/skills"],"capabilities":["agents/skills/agent-configs"]},"lang":"Shell","license":"MIT","maturity":"unreleased","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:alibaba/open-code-review","kind":"repo","title":"open-code-review","why":"Hybrid code review: deterministic pipelines plus an LLM agent, with line-level comments and security rules.","keys":{"races":["cyber/appsec","dev-tools/code-review"],"capabilities":["cyber/appsec/static-analysis","dev-tools/code-review/ai-code-review"]},"lang":"Go","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:alicanakyol/sems","kind":"repo","title":"sems","why":"Detects VMs, sandboxes and analysis tools (VirtualBox, Cuckoo, QEMU, Sandboxie) from inside a process.","keys":{"races":["cyber/offensive"],"capabilities":["cyber/offensive/sandbox-evasion"]},"lang":"C++","license":"none","maturity":"quiet","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:alishahryar1/free-claude-code","kind":"repo","title":"free-claude-code","why":"Runs Claude Code, Codex, OpenCode and other harnesses against free model providers.","keys":{"races":["llm-infra/gateways"],"capabilities":["llm-infra/gateways/api-proxy","llm-infra/gateways/free-providers"]},"lang":"Python","license":"NOASSERTION","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:anomalyco/opencode","kind":"repo","title":"opencode","why":"Open-source terminal coding agent that works with any model provider.","keys":{"races":["agents/coding"],"capabilities":["agents/coding/cli-agents"]},"lang":"TypeScript","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:apple/container","kind":"repo","title":"container","why":"Apple's Swift tool for running Linux containers in lightweight VMs on Apple silicon.","keys":{"races":["infra/containers"],"capabilities":["infra/containers/macos-containers"]},"lang":"Swift","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:awesomedata/awesome-public-datasets","kind":"repo","title":"awesome-public-datasets","why":"Topic-sorted list of high-quality open datasets.","keys":{"races":["data/datasets"],"capabilities":["data/datasets/dataset-lists"]},"lang":null,"license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:bytedance/deer-flow","kind":"repo","title":"deer-flow","why":"Long-horizon agent harness that researches, codes and creates, with sandboxes, memory, skills and subagents.","keys":{"races":["agents/assistants"],"capabilities":["agents/assistants/deep-research"]},"lang":"Python","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:can1357/oh-my-pi","kind":"repo","title":"oh-my-pi","why":"pi-based coding agent with the IDE wired in; multi-provider.","keys":{"races":["agents/coding"],"capabilities":["agents/coding/cli-agents"]},"lang":"TypeScript","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:cathrynlavery/diagram-design","kind":"repo","title":"diagram-design","why":"Agent skill for editorial diagrams: 44 types as self-contained HTML and SVG.","keys":{"races":["dev-tools/diagrams"],"capabilities":["dev-tools/diagrams/diagram-skills"]},"lang":"HTML","license":"MIT","maturity":"unreleased","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:containers/bubblewrap","kind":"repo","title":"bubblewrap","why":"Unprivileged Linux sandboxing with user namespaces; the sandbox behind Flatpak.","keys":{"races":["infra/isolation"],"capabilities":["infra/isolation/unprivileged-sandboxes"]},"lang":"C","license":"NOASSERTION","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:coreclaw/coreclaw","kind":"repo","title":"coreclaw","why":"Lightweight TypeScript bot framework with explicit delivery guarantees and failure recovery.","keys":{"races":["agents/frameworks"],"capabilities":["agents/frameworks/bot-frameworks"]},"lang":"TypeScript","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:cporter202/coreclaw-api-directory","kind":"repo","title":"coreclaw-api-directory","why":"Categorized directory of 118 CoreClaw Worker APIs for scraping and automation.","keys":{"races":["agents/tools"],"capabilities":["agents/tools/api-directories"]},"lang":"Python","license":"none","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:cursor/plugins","kind":"repo","title":"plugins","why":"Cursor's plugin specification and official plugins.","keys":{"races":["agents/skills"],"capabilities":["agents/skills/plugin-specs"]},"lang":"TypeScript","license":"none","maturity":"unreleased","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:danyuchn/asd-ste100-skill","kind":"repo","title":"asd-ste100-skill","why":"Skill that rewrites agent-facing English into ASD-STE100 Simplified Technical English.","keys":{"races":["agents/skills"],"capabilities":["agents/skills/writing-skills"]},"lang":"Python","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:debpalash/voicestudio","kind":"repo","title":"VoiceStudio","why":"Fully local ElevenLabs alternative: voice cloning, voice design, dubbing and transcription.","keys":{"races":["media/speech"],"capabilities":["media/speech/text-to-speech","media/speech/voice-cloning"]},"lang":"Python","license":"AGPL-3.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:deepseek-ai/deepseek-harness","kind":"repo","title":"deepseek-harness","why":"DeepSeek's coding-agent harness where everything is a plugin, built on Cordis.","keys":{"races":["agents/coding"],"capabilities":["agents/coding/cli-agents","agents/coding/plugin-architectures"]},"lang":"TypeScript","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:deusdata/codebase-memory-mcp","kind":"repo","title":"codebase-memory-mcp","why":"MCP server that indexes codebases into a persistent knowledge graph; one static C binary.","keys":{"races":["agents/code-intel"],"capabilities":["agents/code-intel/code-graphs"]},"lang":"C","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:dietrichgebert/ponytail","kind":"repo","title":"ponytail","why":"Skill that makes coding agents write the least code: YAGNI, reuse and the stdlib first.","keys":{"races":["agents/skills"],"capabilities":["agents/skills/engineering-skills"]},"lang":"JavaScript","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:earthtojake/text-to-cad","kind":"repo","title":"text-to-cad","why":"Gives agents CAD skills: generate STEP and STL parts from text.","keys":{"races":["agents/tools"],"capabilities":["agents/tools/cad"]},"lang":"Python","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:garrytan/gbrain","kind":"repo","title":"gbrain","why":"Opinionated knowledge brain for OpenClaw and Hermes agents, with schema packs.","keys":{"races":["agents/memory"],"capabilities":["agents/memory/knowledge-brains"]},"lang":"TypeScript","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:gepa-ai/gepa","kind":"repo","title":"gepa","why":"Reflective optimizer that evolves prompts and code from execution traces.","keys":{"races":["agents/optimization"],"capabilities":["agents/optimization/prompt-optimization"]},"lang":"Jupyter Notebook","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:google-research/harness2","kind":"repo","title":"harness2","why":"Research framework where agents recursively refine their own harness on unseen tasks.","keys":{"races":["agents/optimization"],"capabilities":["agents/optimization/harness-optimization"]},"lang":"Python","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:google/agentmesh","kind":"repo","title":"agentmesh","why":"Peer-to-peer mesh network for agents over A2A and MCP.","keys":{"races":["agents/orchestration"],"capabilities":["agents/orchestration/agent-networks"]},"lang":"Go","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:google/gvisor","kind":"repo","title":"gvisor","why":"Application kernel that sandboxes containers by implementing Linux syscalls in user space.","keys":{"races":["infra/isolation"],"capabilities":["infra/isolation/application-kernels"]},"lang":"Go","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:greydgl/pentestgpt","kind":"repo","title":"PentestGPT","why":"LLM-driven agentic framework for automated penetration testing.","keys":{"races":["cyber/offensive"],"capabilities":["cyber/offensive/ai-pentesting"]},"lang":"Python","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:guymanndude/mnemo-cortex","kind":"repo","title":"mnemo-cortex","why":"Local memory coprocessor for agents: persistent recall, semantic search and verified facts, as an MCP server.","keys":{"races":["agents/memory"],"capabilities":["agents/memory/long-term-memory"]},"lang":"Python","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:headroomlabs-ai/headroom","kind":"repo","title":"headroom","why":"Compresses tool outputs, logs and RAG chunks before they reach the LLM; library, proxy or MCP server.","keys":{"races":["agents/context"],"capabilities":["agents/context/output-compression"]},"lang":"Python","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:hjanuschka/pi-multi-pass","kind":"repo","title":"pi-multi-pass","why":"pi extension for using several OAuth subscription accounts per provider.","keys":{"races":["llm-infra/gateways"],"capabilities":["llm-infra/gateways/multi-account"]},"lang":"TypeScript","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:infisical/infisical","kind":"repo","title":"infisical","why":"Open-source platform for secrets, certificates and PKI, with access controls.","keys":{"races":["cyber/identity"],"capabilities":["cyber/identity/secrets-management"]},"lang":"TypeScript","license":"NOASSERTION","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:kaliiiiiiiiii-vinyzu/patchright","kind":"repo","title":"patchright","why":"Undetected drop-in Playwright that avoids bot detection.","keys":{"races":["agents/tools"],"capabilities":["agents/tools/browser-automation"]},"lang":"TypeScript","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:keycloak/keycloak","kind":"repo","title":"keycloak","why":"Open-source identity and access management with OIDC, SAML and SSO.","keys":{"races":["cyber/identity"],"capabilities":["cyber/identity/iam"]},"lang":"Java","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:kkkkhazix/aihot","kind":"repo","title":"AIHOT","why":"Self-hosted site framework that finds trending topics and writes daily digests from your own sources; Chinese-language.","keys":{"races":["media/news"],"capabilities":["media/news/news-digests"]},"lang":"TypeScript","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:krishhgg/sys1rust","kind":"repo","title":"sys1rust","why":"Rust server with custom Metal kernels for fast Laya System 1 inference on Apple silicon.","keys":{"races":["llm-infra/inference"],"capabilities":["llm-infra/inference/apple-silicon"]},"lang":"Rust","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:krishhgg/t3term","kind":"repo","title":"t3term","why":"Terminal client for T3 Code: one Rust binary, interactive or scriptable.","keys":{"races":["agents/clients"],"capabilities":["agents/clients/terminal-clients"]},"lang":"Rust","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:krishhgg/tokenstash","kind":"repo","title":"tokenstash","why":"Writes API keys into .env.local for agents without the keys ever passing through the chat.","keys":{"races":["cyber/agent-security"],"capabilities":["cyber/agent-security/secret-isolation"]},"lang":"Rust","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:kvcache-ai/agentenv","kind":"repo","title":"AgentENV","why":"Distributed platform for running agent environments at scale, E2B-compatible.","keys":{"races":["agents/sandboxes"],"capabilities":["agents/sandboxes/sandbox-fleets"]},"lang":"Rust","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:langchain-ai/openwiki","kind":"repo","title":"openwiki","why":"CLI that writes and maintains agent-facing documentation for a codebase.","keys":{"races":["agents/code-intel"],"capabilities":["agents/code-intel/codebase-docs"]},"lang":"TypeScript","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:microsoft/bitnet","kind":"repo","title":"BitNet","why":"Official inference framework for 1-bit LLMs (BitNet b1.58).","keys":{"races":["llm-infra/inference"],"capabilities":["llm-infra/inference/low-bit"]},"lang":"C++","license":"MIT","maturity":"unreleased","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:mindmuxai/brain.md","kind":"repo","title":"brain.md","why":"File-based project memory for coding agents: decisions, requirements and constraints in markdown.","keys":{"races":["agents/memory"],"capabilities":["agents/memory/project-memory"]},"lang":"JavaScript","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:mksglu/context-mode","kind":"repo","title":"context-mode","why":"Shrinks agent context: sandboxes tool output, persists session memory and routes through MCP and hooks.","keys":{"races":["agents/context"],"capabilities":["agents/context/output-compression"]},"lang":"TypeScript","license":"NOASSERTION","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:moerizzy/tofmapper","kind":"repo","title":"TOFMapper","why":"Semantic segmentation of trees outside forests in high-resolution aerial images.","keys":{"races":["ml/vision"],"capabilities":["ml/vision/aerial-segmentation"]},"lang":"Python","license":"GPL-3.0","maturity":"unreleased","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:moonrepo/moon","kind":"repo","title":"moon","why":"Rust build system and monorepo task runner for JavaScript, TypeScript and more.","keys":{"races":["dev-tools/build"],"capabilities":["dev-tools/build/monorepo-tools"]},"lang":"Rust","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:morluto/rea","kind":"repo","title":"rea","why":"Agent-driven reverse engineering from app behavior down to native binaries, with Ghidra and Hopper.","keys":{"races":["cyber/reversing"],"capabilities":["cyber/reversing/agentic-re"]},"lang":"TypeScript","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:mvschwarz/openrig","kind":"repo","title":"openrig","why":"Persistent agent teams from Claude Code, Codex and Pi, with roles, shared context and owned work.","keys":{"races":["agents/orchestration"],"capabilities":["agents/orchestration/agent-teams"]},"lang":"TypeScript","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:nationalsecurityagency/ghidra","kind":"repo","title":"ghidra","why":"NSA's software reverse engineering framework: disassembler, decompiler and scripting.","keys":{"races":["cyber/reversing"],"capabilities":["cyber/reversing/disassemblers"]},"lang":"Java","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:nousresearch/hermes-agent","kind":"repo","title":"hermes-agent","why":"Nous Research's self-improving personal agent that grows skills and memory as you use it.","keys":{"races":["agents/assistants"],"capabilities":["agents/assistants/personal-agents"]},"lang":"Python","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:nousresearch/hermes-agent-self-evolution","kind":"repo","title":"hermes-agent-self-evolution","why":"Evolves Hermes Agent's skills, prompts and code with DSPy and GEPA.","keys":{"races":["agents/optimization"],"capabilities":["agents/optimization/prompt-optimization"]},"lang":"Python","license":"none","maturity":"unreleased","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:nvidia-nemo/switchyard","kind":"repo","title":"Switchyard","why":"Routes LLM traffic across models and providers while keeping native OpenAI and Anthropic APIs.","keys":{"races":["llm-infra/gateways"],"capabilities":["llm-infra/gateways/llm-gateway","llm-infra/gateways/model-routing"]},"lang":"Rust","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:nvidia/cuvs","kind":"repo","title":"cuvs","why":"GPU library for vector search and clustering, on CUDA.","keys":{"races":["llm-infra/rag"],"capabilities":["llm-infra/rag/vector-search"]},"lang":"Cuda","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:nvidia/nemo-fabric","kind":"repo","title":"NeMo-Fabric","why":"One interface for configuring, running and comparing agent harnesses, through per-harness adapters.","keys":{"races":["agents/orchestration"],"capabilities":["agents/orchestration/harness-adapters"]},"lang":"Python","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:nvidia/nemoclaw","kind":"repo","title":"NemoClaw","why":"Runs OpenClaw, Hermes and Deep Agents inside NVIDIA OpenShell with managed inference.","keys":{"races":["agents/sandboxes"],"capabilities":["agents/sandboxes/secure-runtimes"]},"lang":"TypeScript","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:nvidia/openshell","kind":"repo","title":"OpenShell","why":"NVIDIA's safe, private runtime for autonomous agents.","keys":{"races":["agents/sandboxes"],"capabilities":["agents/sandboxes/secure-runtimes"]},"lang":"Rust","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:nvidia/personal-ai-router","kind":"repo","title":"Personal-AI-Router","why":"Distributes inference across the devices on a home network.","keys":{"races":["llm-infra/inference"],"capabilities":["llm-infra/inference/distributed-inference"]},"lang":"Go","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:nvidia/skillspector","kind":"repo","title":"SkillSpector","why":"Scans agent skills for prompt injection, data exfiltration and supply-chain risks before install.","keys":{"races":["cyber/agent-security"],"capabilities":["cyber/agent-security/skill-scanning"]},"lang":"Python","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:obneyai/grain","kind":"repo","title":"grain","why":"Clojure framework for event-sourced applications built through natural-language programming.","keys":{"races":["agents/frameworks"],"capabilities":["agents/frameworks/agent-workflows"]},"lang":"Clojure","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:obneyai/orc","kind":"repo","title":"orc","why":"Clojure lab for building accountable agentic features, with behavior trees for AI workflows.","keys":{"races":["agents/frameworks"],"capabilities":["agents/frameworks/agent-workflows"]},"lang":"Clojure","license":"MIT","maturity":"unreleased","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:obneyai/sio","kind":"repo","title":"sio","why":"Clojure library for structured LLM input and output, with forgiving parsing of nested data.","keys":{"races":["agents/frameworks"],"capabilities":["agents/frameworks/structured-io"]},"lang":"Clojure","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:obsidianmd/pele","kind":"repo","title":"pele","why":"Fast Mermaid diagram renderer from the Obsidian team.","keys":{"races":["dev-tools/diagrams"],"capabilities":["dev-tools/diagrams/mermaid-rendering"]},"lang":"TypeScript","license":"NOASSERTION","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:onyx-dot-app/onyx","kind":"repo","title":"onyx","why":"Self-hosted AI chat and enterprise search over your documents, with any LLM.","keys":{"races":["llm-infra/chat-apps","llm-infra/rag"],"capabilities":["llm-infra/chat-apps/self-hosted-chat","llm-infra/rag/knowledge-platforms"]},"lang":"Python","license":"NOASSERTION","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:ovexro/dockpanel","kind":"repo","title":"dockpanel","why":"Server control panel in Rust and React: sites, databases, Docker apps, mail, DNS and backups.","keys":{"races":["infra/hosting"],"capabilities":["infra/hosting/control-panels"]},"lang":"Rust","license":"AGPL-3.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:pain2hustle/amh-cloudflare-ops-mcp-by-wt","kind":"repo","title":"AMH-Cloudflare-Ops-MCP-by-WT","why":"Approval-gated MCP for inspecting, publishing and verifying Cloudflare site changes.","keys":{"races":["agents/tools"],"capabilities":["agents/tools/cloud-ops"]},"lang":"HTML","license":"NOASSERTION","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:panniantong/agent-reach","kind":"repo","title":"Agent-Reach","why":"One CLI for agents to read and search Twitter, Reddit, YouTube, GitHub and more, with no API fees.","keys":{"races":["agents/tools"],"capabilities":["agents/tools/web-access"]},"lang":"Python","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:paperclipai/paperclip","kind":"repo","title":"paperclip","why":"Open-source app for managing teams of agents at work: orchestration, governance and workflows.","keys":{"races":["agents/orchestration"],"capabilities":["agents/orchestration/agent-teams"]},"lang":"TypeScript","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:peterldowns/pgtestdb","kind":"repo","title":"pgtestdb","why":"Gives each Go test its own temporary, isolated Postgres database, fast via template databases.","keys":{"races":["dev-tools/testing"],"capabilities":["dev-tools/testing/test-databases"]},"lang":"Go","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:pingdotgg/t3code","kind":"repo","title":"t3code","why":"Control surface for the coding agents on your machine (Claude Code, Codex, Cursor and more), with a mobile app.","keys":{"races":["agents/clients"],"capabilities":["agents/clients/desktop-apps","agents/clients/mobile-apps"]},"lang":"TypeScript","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:primeintellect-ai/prime-agent","kind":"repo","title":"prime-agent","why":"Self-improving coding agent built for long-running autonomous tasks.","keys":{"races":["agents/coding"],"capabilities":["agents/coding/cli-agents","agents/coding/long-running"]},"lang":"Rust","license":"NOASSERTION","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:primeintellect-ai/verifiers","kind":"repo","title":"verifiers","why":"Library for building RL environments and evals for LLMs.","keys":{"races":["agents/optimization"],"capabilities":["agents/optimization/rl-environments"]},"lang":"Python","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:protocolbuffers/protobuf","kind":"repo","title":"protobuf","why":"Google's schema-based serialization format and its compiler.","keys":{"races":["data/serialization"],"capabilities":["data/serialization/schema-formats"]},"lang":"C++","license":"NOASSERTION","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:pwndbg/pwndbg","kind":"repo","title":"pwndbg","why":"GDB and LLDB plugin for exploit development and reverse engineering.","keys":{"races":["cyber/reversing"],"capabilities":["cyber/reversing/debuggers"]},"lang":"Python","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:pydantic/monty","kind":"repo","title":"monty","why":"Secure Python sandbox written in Rust for running AI-written code.","keys":{"races":["agents/sandboxes"],"capabilities":["agents/sandboxes/code-sandboxes"]},"lang":"Rust","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:ranxianglei/billion-context","kind":"repo","title":"billion-context","why":"Context-compression plugin for month-long sessions in small context windows.","keys":{"races":["agents/context"],"capabilities":["agents/context/context-compaction"]},"lang":"TypeScript","license":"NOASSERTION","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:router-for-me/cliproxyapi","kind":"repo","title":"CLIProxyAPI","why":"OpenAI/Claude/Gemini-compatible gateway in Go over CLI subscriptions.","keys":{"races":["llm-infra/gateways"],"capabilities":["llm-infra/gateways/api-proxy","llm-infra/gateways/llm-gateway"]},"lang":"Go","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:rtk-ai/rtk","kind":"repo","title":"rtk","why":"CLI proxy that cuts LLM tokens on common dev commands by 60-90%; one Rust binary.","keys":{"races":["agents/context"],"capabilities":["agents/context/output-compression"]},"lang":"Rust","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:runvendo/vendo","kind":"repo","title":"vendo","why":"SDK for embedded agents that your customers use inside your product.","keys":{"races":["agents/frameworks"],"capabilities":["agents/frameworks/embedded-agents"]},"lang":"TypeScript","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:scgopi/graphcode","kind":"repo","title":"GraphCode","why":"macOS app for orchestrating parallel coding agents as a graph.","keys":{"races":["agents/clients","agents/orchestration"],"capabilities":["agents/clients/desktop-apps","agents/orchestration/agent-teams"]},"lang":"Swift","license":"NOASSERTION","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:seaweedfs/seaweedfs","kind":"repo","title":"seaweedfs","why":"Distributed storage for S3 objects, files and Iceberg tables, with O(1) disk access.","keys":{"races":["data/storage"],"capabilities":["data/storage/object-storage"]},"lang":"Go","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:secureagentics/adrian","kind":"repo","title":"Adrian","why":"Runtime monitor that catches malicious tool use, prompt injection and policy drift before an agent acts.","keys":{"races":["cyber/agent-security"],"capabilities":["cyber/agent-security/runtime-guardrails"]},"lang":"Python","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:shepherd-agents/shepherd","kind":"repo","title":"shepherd","why":"Records agent runs as Git-like traces so meta-agents can fork, replay and revert them.","keys":{"races":["agents/sandboxes"],"capabilities":["agents/sandboxes/run-replay"]},"lang":"Python","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:steipete/agent-scripts","kind":"repo","title":"agent-scripts","why":"Shared agent scripts, skills and instructions, synced across repositories.","keys":{"races":["agents/skills"],"capabilities":["agents/skills/agent-configs"]},"lang":"Shell","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:tashfeenahmed/freellmapi","kind":"repo","title":"freellmapi","why":"One OpenAI-compatible endpoint over 34 free LLM providers, with smart routing and failover.","keys":{"races":["llm-infra/gateways"],"capabilities":["llm-infra/gateways/free-providers","llm-infra/gateways/llm-gateway","llm-infra/gateways/model-routing"]},"lang":"TypeScript","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:tencent/weknora","kind":"repo","title":"WeKnora","why":"LLM knowledge platform: documents become a queryable RAG, a reasoning agent and a self-maintaining wiki.","keys":{"races":["llm-infra/rag"],"capabilities":["llm-infra/rag/knowledge-platforms"]},"lang":"Go","license":"NOASSERTION","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:tencentcloud/octop","kind":"repo","title":"Octop","why":"Self-hosted multi-user, multi-agent assistant with long-term memory.","keys":{"races":["agents/assistants"],"capabilities":["agents/assistants/team-assistants"]},"lang":"Python","license":"MIT","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:testcontainers/testcontainers-go","kind":"repo","title":"testcontainers-go","why":"Go library that runs throwaway Docker containers as test dependencies.","keys":{"races":["dev-tools/testing"],"capabilities":["dev-tools/testing/integration-testing"]},"lang":"Go","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:tester-army/e2e","kind":"repo","title":"e2e","why":"End-to-end testing framework for web and mobile apps.","keys":{"races":["dev-tools/testing"],"capabilities":["dev-tools/testing/e2e-testing"]},"lang":"TypeScript","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:thedotmack/claude-mem","kind":"repo","title":"claude-mem","why":"Captures what an agent does, compresses it, and injects relevant context into later sessions.","keys":{"races":["agents/memory"],"capabilities":["agents/memory/long-term-memory"]},"lang":"TypeScript","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:thisisbankole/handshake","kind":"repo","title":"handshake","why":"Carries session context from one AI coding agent to another.","keys":{"races":["agents/memory"],"capabilities":["agents/memory/session-handoff"]},"lang":"Go","license":"none","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:unxsist/jet-pilot","kind":"repo","title":"jet-pilot","why":"Native Kubernetes desktop client for every cluster and cloud account, with no telemetry.","keys":{"races":["infra/kubernetes"],"capabilities":["infra/kubernetes/desktop-clients"]},"lang":"TypeScript","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:usestrix/strix","kind":"repo","title":"strix","why":"Open-source AI agents that pentest your app and help fix what they find.","keys":{"races":["cyber/appsec","cyber/offensive"],"capabilities":["cyber/appsec/vuln-discovery","cyber/offensive/ai-pentesting"]},"lang":"Python","license":"Apache-2.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:valstorm-corp/valstorm-agent","kind":"repo","title":"valstorm-agent","why":"ReAct agent runtime for workstations, CI pipelines and team automation.","keys":{"races":["agents/assistants"],"capabilities":["agents/assistants/task-automation"]},"lang":"Python","license":"NOASSERTION","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:vectorize-io/hindsight","kind":"repo","title":"hindsight","why":"Long-term agent memory that learns from experience.","keys":{"races":["agents/memory"],"capabilities":["agents/memory/long-term-memory"]},"lang":"Python","license":"MIT","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:vercel/ai","kind":"repo","title":"ai","why":"Provider-agnostic TypeScript SDK for building LLM apps and agents.","keys":{"races":["agents/frameworks"],"capabilities":["agents/frameworks/llm-sdks"]},"lang":"TypeScript","license":"NOASSERTION","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:visa/visa-vulnerability-agentic-harness","kind":"repo","title":"visa-vulnerability-agentic-harness","why":"Visa's pipeline that finds, verifies and fixes source-code vulnerabilities with deterministic and agentic stages.","keys":{"races":["cyber/appsec"],"capabilities":["cyber/appsec/auto-remediation","cyber/appsec/vuln-discovery"]},"lang":"Python","license":"NOASSERTION","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:volcengine/openviking","kind":"repo","title":"OpenViking","why":"Context database for agents that unifies memory, RAG and skills.","keys":{"races":["agents/memory"],"capabilities":["agents/memory/context-databases"]},"lang":"Python","license":"AGPL-3.0","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:xai-org/grok-build","kind":"repo","title":"grok-build","why":"Grok's coding-agent harness with a fullscreen, mouse-driven TUI.","keys":{"races":["agents/coding"],"capabilities":["agents/coding/cli-agents"]},"lang":"Rust","license":"Apache-2.0","maturity":"new","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}},
{"id":"gh:yugabyte/yugabyte-db","kind":"repo","title":"yugabyte-db","why":"Cloud-native distributed SQL database, PostgreSQL-compatible.","keys":{"races":["data/databases"],"capabilities":["data/databases/distributed-sql"]},"lang":"C","license":"NOASSERTION","maturity":"stable","status":null,"stats":{"pickers":0,"packs":1,"used_30d":0,"star_delta_7d":0}}
]}