Managed Projects

AgentsKit

  Analyzed about 18 hours ago

AgentsKit is a free, MIT-licensed TypeScript ecosystem for building production AI agents without locking an application to one model provider, framework, database, or hosted platform. Stable contracts connect models, runtimes, tools, skills, memory, RAG, observability, sandboxing, evaluation, and ... [More] interfaces, so teams can replace any layer without rewriting the product. The open ecosystem includes Chat, Registry, Agents Playbook, Doc Bridge, and Code Review, and supports hosted or local models and self-hosting. AKOS is separate and optional: an enterprise control plane for organizations that need governed AI operations. [Less]

123K lines of code

0 current contributors

1 day since last commit

0 users on Open Hub

New Project
0.0
 
I Use This

doc-bridge

  Analyzed 1 day ago

A documentation boundary for coding agents. Doc Bridge converts repository docs into deterministic, read-only handoffs that identify what to read first, where changes are allowed, who owns the area, and which checks must pass. It runs as a CLI or MCP server and integrates with Fumadocs and ... [More] Docusaurus, helping teams replace guesswork with auditable repository context. [Less]

13.3K lines of code

0 current contributors

3 days since last commit

0 users on Open Hub

New Project
0.0
 
I Use This

AgentsKit Chat

  Analyzed 1 day ago

A cross-framework application framework for interactive AI agent experiences. AgentsKit Chat provides provider-neutral streaming, headless UI primitives, conversation state, tool interactions, and adapters without forcing teams into a hosted platform or model vendor. It is designed to fit existing ... [More] JavaScript applications and keep the interaction layer replaceable. [Less]

16.5K lines of code

0 current contributors

4 days since last commit

0 users on Open Hub

New Project
0.0
 
I Use This

AgentsKit Registry

  Analyzed 1 day ago

A source-first registry of ready-to-use AI agents. Instead of installing opaque hosted agents, developers copy the implementation into their own codebase, inspect every decision, customize behavior, and keep full control of providers, tools, data, and deployment. It brings a shadcn-style ownership model to reusable agent capabilities.

33.1K lines of code

0 current contributors

14 days since last commit

0 users on Open Hub

New Project
0.0
 
I Use This

AgentsKit Playbook

  Analyzed 1 day ago

An open, executable playbook for shipping production software with AI coding agents. It turns engineering expectations into reusable pillars, patterns, prompts, and CI-enforceable gates so teams can scale agent-assisted development without relying on unwritten rules or a specific coding assistant.

3.23K lines of code

0 current contributors

13 days since last commit

0 users on Open Hub

New Project
0.0
 
I Use This

AgentsKit Code Review

  Analyzed 1 day ago

A low-noise AI code review tool available as a CLI and GitHub Action. It analyzes pull requests with the open AgentsKit code-review agent, emphasizes actionable findings, and keeps the review workflow portable across providers and repositories instead of tying teams to a proprietary review service.

1.25K lines of code

0 current contributors

4 days since last commit

0 users on Open Hub

New Project
0.0
 
I Use This