◆ 282 能力 · 280 scanned & graded
🛡 280 scanned & graded · 41% rated safe · see the findings →经过安全审核的 AI 能力注册中心
每一款技能、MCP server、agent、数据集与评测都经过审计，安全性以 0–100 分评级，并持续追踪更新时效——让人类和 agent 都能放心采用真正信得过的 AI 能力。
◆ Free security scanner
Scan any AI tool — free
Paste any public GitHub repo and get an instant security grade — malware, embedded secrets, dangerous code, known CVEs, and prompt-injection surface. Free with a free account. Clean, open-source repos automatically join the catalog, graded and searchable.
为什么选择 ai-supply
放心采用真正值得信任的 AI 能力
◈
我们会审读代码
我们提取每个能力的源代码，并按双轴模型评分——只有恶意或被篡改的代码才会降低评分，因此那些因用途本身而具有风险的工具不会被错误地埋没。
⚑
安装前先了解其暴露面
每个条目都附带一份能力清单（capability manifest）——列出其工具、权限（文件系统、Shell、网络、密钥）以及出站主机（egress）——让你清楚看到自己授予了什么。
⌘
免费，且为智能体而生
智能体在安装前会核查能力的信任评定，并执行你设定的策略。一切均免费且开放。
◆ 专为智能体设计
人可发现——智能体完全可操作
agent 与人类享有完全对等的能力：只需一个限定权限的 API key，即可通过 JSON API 或原生 MCP 工具进行搜索、安装、下载、评价与发布。将 Claude Code、Cursor 或任何支持函数调用的 agent 指向 ai-supply，让它从人类所用的同一份安全审核目录中自助取用。
claude mcp add ai-supply --env AIM_API_KEY=YOUR_KEY -- npx -y ai-supply-mcp
Vetted & graded
Most secure capabilities
⬡流水线
Prompt flow
Microsoft's open-source toolkit for building, evaluating, and deploying high-quality LLM applications end-to-end.
ai-supply
✓ A · 100⟳ 1y agosecretsshellnetwork
↓ 167k
⬡流水线
PyBullet / Bullet3
Real-time physics simulation for robotics and game AI using the Bullet physics engine with Python bindings.
ai-supply
✓ A · 100⟳ 4y agosecretsfs
↓ 76k
⬡流水线
PyTorch3D — 3D Deep Learning Library
Meta FAIR's PyTorch library for deep learning with 3D meshes, point clouds, and volumetric data.
ai-supply
✓ A · 100⟳ 8mo agosecretsshellnetwork
↓ 136k
◆技能
Argilla — Collaborative Data Annotation for LLMs
Open-source annotation platform for building high-quality fine-tuning and RLHF datasets; integrates with Hugging Face Hub.
ai-supply
✓ A · 100⟳ 1y agonetworkfs
↓ 38k
◆技能
Ponytail
Makes your AI agent think like the laziest senior dev — the best code is the code you never wrote (YAGNI-first).
ai-supply
✓ A · 100⟳ 29d agoshellnetworkfs
↓ 30k
⇄连接器
DuckDB
In-process OLAP SQL database that runs inside your Python, R, or Node.js process — no server required.
ai-supply
✓ A · 100⟳ 6d agonetworkfs
↓ 597k
浏览
分类
CybersecurityThreat detection, fuzzing, red-team agents.CodingCode synthesis, refactoring, repo-aware agents.FinanceTrading, risk, accounting, on-chain analysis.Agentic capabilityPlanners, tool-use, browser, computer-use.MarketingCopy, ads, segmentation, growth ops.OrchestrationMulti-agent graphs, queues, schedulers.Data & ETLConnectors, pipelines, vector stores.ResearchWeb, papers, knowledge graphs, citation.Vision & ImageGeneration, detection, OCR, segmentation.Audio & SpeechTTS, STT, music, diarization.Language & NLPTranslation, summarization, extraction.DevOps & InfraCI/CD, IaC, observability, incident response.Robotics & ControlMotion, perception, simulation, control.HealthcareClinical, biomedical, imaging assistants.Legal & ComplianceContracts, policy, eDiscovery, audit.Gaming & SimulationNPCs, procedural content, balancing.
编辑精选
精选推荐
⊕插件
LangChain
The leading framework for building LLM-powered applications and agents with chains, tools, memory, and retrieval.
ai-supply
! B · 75⟳ 5d agoshellnetworkfs
↓ 234k
◇MCP 服务器
MCP Reference Servers
Official reference MCP servers: filesystem, git, fetch, memory, and more — the fastest way to give agents real-world tool access.
ai-supply
! B · 88⟳ 18d agoshellnetworkfs
↓ 215k
◉智能体
Browser Use
Connect AI agents to the web: Browser Use lets LLMs browse, interact, and extract from any website autonomously.
ai-supply
✓ A · 100⟳ 11d agoshellinstall-hooknetwork
↓ 164k
最新发布
最新审核
⇄连接器
LLM (CLI)
CLI and Python library to access many LLM providers with a plugin system and SQLite logging.
ai-supply
✓ A · 100⟳ 19d agoshellnetworkfs
↓ 47k
⌬工作流
Marvin
Prefect's Python library for building reliable AI functions, tasks, and workflow orchestration.
ai-supply
! B · 75⟳ 4mo agosecretsshellnetwork
↓ 7.8k
⌬工作流
ComfyUI
Modular node-graph GUI, API, and backend for diffusion-model image/media generation pipelines.
ai-supply
! B · 75⟳ 13d agosecretsshellnetwork
↓ 35k
▣数据集
HumanEval
OpenAI's 164-problem Python benchmark for evaluating code-generation correctness (pass@k).
ai-supply
! B · 88⟳ 1y ago
↓ 3.6k
▣数据集
MMLU Benchmark
MMLU: 57-subject multiple-choice benchmark for broad LLM knowledge and reasoning.
ai-supply
✓ A · 100⟳ 3y ago
↓ 44k
▣数据集
Anthropic HH-RLHF Dataset
Anthropic's human-preference (helpful/harmless) dataset for RLHF and alignment research.
ai-supply
✓ A · 100⟳ 1y agoPII
↓ 51k