SuperAPI AI Directory
Partner with us中文AI newsFavorites

AI developer platforms

3,631 results · Page 75 of 101

hcom

AI developer platforms

hcom is a CLI tool that lets multiple AI coding agents message each other, observe status, and spawn new agents across different terminals. It supports many AI command-line tools such as claude, codex, opencode, copilot, qoder, grok, pi, omp, agy, cursor, kimi, kilo, gemini, and more. To use it, prefix the command that starts an agent with hcom, then prompt normally. It can be used to coordinate multi-agent pipelines, run different AI CLIs as each other's subagents, or avoid manual copy-pasting. All agents run in visible real terminals that can be scrolled and interrupted. Their activity is recorded to a local SQLite database via hooks, and messages are delivered based on that; hooks are activated only when an agent is started with hcom, so normal use is unaffected.

Read more

hector

AI developer platforms

Hector is an open-source AI agent runtime platform written in Go, emphasizing autonomy and control with zero dependencies. It is delivered as a single binary of about 30MB, with a built-in visual Web UI, and can run with just one YAML configuration file, supporting deployment locally, on isolated networks, or in any cloud environment. The platform implements agent interoperability based on the A2A protocol, supports MCP tool connections, and can switch between LLM providers such as Anthropic, OpenAI, Gemini, and Ollama with a single line of configuration. It defines agents, orchestration patterns, RAG pipelines, guardrails, triggers, and notifications in a configuration-driven manner, and provides enterprise-grade security capabilities such as SQL persistence (SQLite/PostgreSQL), multi-tenant isolation, JWKS/OIDC authentication, prompt injection detection, PII redaction, and tool sandboxing. Hector is licensed under MIT and is suitable for teams that need complete control over their AI infrastructure.

Read more

Honeycomb

AI developer platforms

Honeycomb is an observability platform for software in the AI era, offering fast queries, unified telemetry, and LLM observability. Engineering teams can use it to trace distributed services, debug non-deterministic AI workflows, and give humans and agents a shared view of what users actually experience. The platform includes capabilities such as agent and LLM observability, distributed tracing, log management and analysis, and telemetry pipelines, supports OpenTelemetry, and is used by Slack, Intercom, Dropbox, and others.

Read more

HongMaCC

AI developer platforms

HongMaCC (self-branded as HongMaCode on its official website) is an AI API relay service platform for domestic developers, focusing on direct domestic access to model services such as Claude Code, Codex, and Gemini. It claims millisecond-level response, supports Alipay and WeChat Pay, and requires no VPN. It is suitable for developers who need to call mainstream overseas AI programming/chat model APIs within the domestic network environment. The specific list of available models, pricing, call limits, stability, and compliance status are not yet provided.

Read more

How AI chatbots like ChatGPT or Bard work

AI developer platforms

This is a visual explainer article from The Guardian that explains how the large language models (LLMs) behind AI chatbots such as ChatGPT and Bard work. Through visualizations, the article explains core concepts such as the basic architecture of LLMs, nodes and connections, token and vector representations, Transformers, and attention mechanisms, and explains how the model predicts the next most likely word based on context. The content is aimed at general readers and does not involve product features or usage steps.

Read more