SuperAPI AI Directory
Partner with us中文AI newsFavorites

AI agents

2,963 results · Page 49 of 83

Graphlit

AI agents

Graphlit is a context and data platform for AI agents, focusing on providing agents with real-time synchronization and semantic search capabilities. According to the official description, it can synchronize data from sources such as Slack, GitHub, and Jira in real time and provide it through a unified API, with no operations required and built-in semantic search. It is suitable for development scenarios that need to connect agents to multi-source external data and build retrieval and context enhancement capabilities.

Read more

Grit

AI agents

Grit is a source code search and modification tool based on the GritQL declarative query language, designed to automate large-scale code migrations, custom lints, and code quality improvements. GritQL allows users to use code snippets directly as queries without needing to learn AST details, and with Rust and query optimization it supports codebases at the tens-of-millions-of-lines scale. It has a built-in module system, can reuse more than 200 standard patterns, supports multiple languages such as JavaScript/TypeScript, Python, JSON, Java, Terraform, Solidity, CSS, Markdown, YAML, Rust, Go, and SQL, and supports automatic fix rules.

Read more

h5i

AI agents

h5i is a web security testing workspace for AI agents, offering two complementary paths: letting AI agents discover web application vulnerabilities through red team testing, and formally verifying the correctness of Rust code with Lean 4. It provides a unified toolset for browser automation, HTTP traffic manipulation, and reconnaissance, and can be used with any web service without migrating frameworks, and supports replaying attack flows in CI/CD to detect security regressions.

Read more

hello-agents

AI agents

"Building Agents from Scratch" is a systematic agent learning tutorial initiated by the Datawhale community, open source and free. The tutorial balances theory and practice, starting from agent concepts, history, and classic paradigms, guiding learners to use low-code platforms and code frameworks, and to build a self-developed framework HelloAgents from scratch based on the OpenAI native API, gradually mastering advanced technologies such as context engineering, memory and retrieval, multi-agent protocols, evaluation, and Agentic RL, and ultimately completing multi-agent application development through comprehensive cases such as an intelligent travel assistant, an automated deep research agent, and a cyber town. The project is aimed at AI developers, software engineers, students, and self-learners with a Python foundation, who need basic concepts of large language models and do not need a deep algorithm or training background.

Read more

hermes-agent

AI agents

Hermes Agent is a self-improving AI agent developed by Nous Research, with a built-in learning loop that can create skills from experience, improve skills through use, proactively persist knowledge, retrieve its own past conversations, and build a deep understanding model of the user across sessions. It can run on a VPS costing as little as $5, a GPU cluster, or serverless infrastructure that costs almost nothing when idle, is not tied to a local computer, and can be interacted with remotely through messaging platforms such as Telegram. Users can freely choose models (Nous Portal, OpenRouter, OpenAI, or their own endpoint, etc.) and switch using the hermes model command without modifying code. It supports Linux, macOS, WSL2, and native Windows (PowerShell), and provides an APT repository for Android/Termux.

Read more

hivemoot

AI agents

hivemoot is an open-source framework for assembling AI agent teams on GitHub, including governance, role definitions, and a CLI. It enables multiple AI agents to work in a repository as real contributors, automatically opening issues, participating in comment discussions, writing code, reviewing PRs, voting on decisions, and delivering, running autonomously around the clock. You define roles for the agents (such as engineer, reviewer, researcher, guard), and the team collaborates using the same Issues, PRs, and CI workflows as humans. It is not a single chat assistant or autocomplete, but an autonomous team that can propose, debate, and vote on what to build. The framework runs entirely on your own hardware, using your own API keys. It provides a terminal CLI tool to view repository status, and supports any AI agent that can interact with GitHub, such as Claude, GPT-4, Gemini, etc. The project also includes a real demo project, Colony, run entirely independently by agents.

Read more