SuperAPI AI Directory
Partner with us中文AI newsFavorites

AI agents

2,963 results · Page 54 of 83

Loopa

AI agents

Loopa is an AI agent for work automation that can think, plan, and execute complex tasks. It can generate presentations, analyze PDFs, create videos, automatically reply to emails, and run multi-agent workflows without manual operation. It supports scenarios such as research automation, financial reporting, content generation, image and video production, and landing page generation, can connect to everyday applications such as Telegram, Discord, and email, runs in a secure sandbox, and protects data privacy. Start using it for free.

Read more

Luffy

AI agents

Luffy is an AI employee that runs in Slack and can connect to the team's existing tools to automatically complete the delivery of reports, dashboards, follow-up tasks, and internal apps. It can run workflows across multiple tools and execute tasks such as outbound sales, lead discovery and nurturing, and booking demos end to end. Luffy has a built-in native CRM, as well as lead, web, and SEO data, with no extra subscription required. Users can assign tasks to Luffy as if assigning responsibilities, and it will keep working in Slack, WhatsApp, or email, citing sources, requesting approval, and reporting the execution steps. Pricing is usage-based, and there is no charge when no work is done.

Read more

LynxPrompt

AI agents

LynxPrompt is a self-hostable, federation-capable platform for managing AI IDE/tool rules and commands. It can convert one set of AI coding rules into the rule files read by 30 tools at once, including AGENTS.md, CLAUDE.md, .cursor/rules/, Copilot instructions, and more, and generates slash commands for six of those tools. The platform offers both a WebUI and a CLI, with a built-in 12-step wizard that generates rules based on a project's tech stack, commands, code style, and boundary conditions; the blueprint feature supports team sharing and variable substitution; the lynxp CLI and VS Code extension can push rule files to every repository. It supports Docker Compose or Helm deployment, comes with PostgreSQL, and configures all functionality through environment variables. Optional AI editing requires your own Anthropic API key and is disabled by default.

Read more

MagiCrew

AI agents

MagiCrew is an enterprise-grade open-source AI Agent platform designed to enable every employee to deploy a dedicated AI digital employee to handle real tasks, not just chat. It can encapsulate ERP/CRM and business knowledge into reusable AI digital employees, achieving build once and deploy company-wide. The platform supports multi-agent collaboration, with an orchestrator Agent scheduling specialized Agents to work in parallel; AI outputs can be directly transformed into deliverable outcomes such as PPTs, dashboards, professional reports, and Excel files. It also provides enterprise-grade governance: three-tier budget control, human approval for high-risk operations, proprietary sandbox isolation, tracking of AI invocation costs by department/user/task, and compatibility with the Anthropic and OpenClaw Skills ecosystems, defining custom skills through a conversational Skill Creator.

Read more

Makedraft

AI agents

Makedraft is an AI agent tool for real project development. Users only need to describe their requirements in chat, and it will directly create a branch and generate code in a real GitHub repository, with no sandbox or export step required. Each branch is automatically deployed as an independent service, and after code is committed it is automatically redeployed. It supports React and Next.js projects, allows Pro and Business plans to switch models freely, and supports binding custom domains. It is suitable for developers who want to skip environment setup and directly obtain deployable code.

Read more

MartinLoop

AI agents

MartinLoop is a task management platform for coding agents, designed to enable coding agents to complete real software work without step-by-step supervision. It provides explicit task contracts for agent work, including goals, scope, constraints, checks, recovery, results, and run history. Run results fall into three states: VERIFIED, STOPPED, or NEEDS REVIEW, where VERIFIED means the configured checks passed but does not constitute a general guarantee of correctness or safety. MartinLoop is not another agent, nor is it an IDE; the coding work is still done by the agent, while it is responsible for keeping tasks focused, bounded, checkable, recoverable, and recordable.

Read more

MaxClaw

AI agents

MaxClaw is an official cloud-hosted AI agent built by MiniMax on the open-source OpenClaw framework, powered by the 229B-parameter MiniMax M2.5 model. It focuses on one-click deployment, requiring no server, Docker, or manual API key management, and can bring an always-on intelligent assistant online within 10 seconds. MaxClaw supports persistent long-term memory of over 200,000 tokens, can remember past conversations, user preferences, and work habits, and supports custom names, personalities, and behavioral styles. It natively integrates with Telegram, Discord, and Slack, and can be invoked in commonly used communication channels with one click. Its toolchain inherits the OpenClaw ecosystem, supporting complex multi-step workflows such as web browsing, code execution, file analysis, automation scripts, and schedule management. The official claim is that it provides intelligence close to frontier models at about 1/7 to 1/20 of the cost, making it suitable for high-frequency automation tasks.

Read more