SuperAPI AI Directory
Partner with us中文AI newsFavorites

AI agents

2,946 results · Page 20 of 82

somora_agent

AI agents

somora is a personal AI agent team server that runs on the user's own machine. It supports configuring multiple agents with independent personalities, private memories, model preferences, and tool permissions, and these agents can define reporting relationships and delegate tasks to each other through team.yaml. Users can switch between Claude, ChatGPT, Grok, or any OpenAI-compatible endpoint (such as Ollama, LM Studio, vLLM, OpenRouter) via slash commands within the same conversation, while history and tools remain unchanged. Agents record facts while chatting, and a background "dreaming" cycle promotes stable knowledge into a shared Obsidian wiki. All data (configuration, agent memories, sessions, attachments) is stored under ~/.somora/. Three clients are provided: terminal, browser, and mobile app, all connecting to the same local server. The project is in active development and welcomes early testers.

Read more

Spawned

AI agents

Spawned is a deployment platform for the AI era, providing a cloud infrastructure interface for AI agents. It allows users to deploy real infrastructure into their own AWS, GCP, or Azure accounts, completing configuration, scaling, and operations in a unified interface. All resources run as real cloud infrastructure, support export and Terraform compatibility from the start, and avoid vendor lock-in. The platform is based on a declarative framework, allowing agents to build in the user's cloud environment while ensuring resources are clear and controllable, reducing issues such as runaway bills, misconfigurations, or forgotten hallucinated components. Deployments, environments, logs, and policy controls are centralized in one operations interface, and resources are grouped by project and fully versioned. If you do not have a cloud account, you can also use its managed version on some clouds, but with limited functionality. Signing up gives you $20 in free compute credits, so you can quickly launch your first project.

Read more

statewave

AI agents

Statewave is an open-source memory runtime for AI agents that replaces query-time retrieval with reproducible, provenance-tagged context packs. It is self-hosted on Postgres + pgvector, provides Python and TypeScript SDKs, is licensed under Apache-2.0, and the current version v1.5.0 is under active development. The core workflow includes ingestion, compilation, retrieval, and governance: recording raw events, compiling them by topic into typed memories with confidence and provenance, assembling ranked context packs with token limits on demand, and enabling auditing of topic timelines, tracing memory provenance, or deleting by topic.

Read more

Stoneforge

AI agents

Stoneforge is an orchestration platform for running multiple AI coding agents in parallel. It avoids merge conflicts by assigning each agent an independent git worktree, and uses a dependency-aware task scheduling daemon to automatically assign ready tasks to idle agents, preventing duplicate work and blocked startups. Users can view each agent's real-time output, kanban-style task progress, merge status, and performance metrics through a real-time Web dashboard. It supports multiple providers such as Claude Code, OpenCode, and OpenAI Codex, and offers capabilities including channels, threads, inbox triage, an evergreen document library, Linear-like issue tracking, resumable structured workflow templates, and automated merge review. The official installation is given as a three-step command, and it claims you can get started in 30 seconds.

Read more