AI Agent2025
Searchable plugin marketplace and curated collections of Claude Code agents, commands, hooks, skills and plugins — lets users discover, browse, and install subagents and tooling via a web UI, plugin commands or a CLI. Indexes community plugins and MCP servers and provides curated packs.
AI Agent2025
Compiles an agent's raw chat logs, documents, and tool traces into three persistent layers — index, learned skills, and user memory — so context survives sessions. Claims 92% Locomo-benchmark accuracy and up to 95% lower token cost than replaying history.
AI Infra2025
Lets AI coding agents provision and operate a full backend themselves — Postgres with pgvector, OAuth2 auth, S3-style storage, Deno edge functions, and hosting — through one interface, plus an OpenAI-compatible model gateway.
AI Agent2025
Evaluates and optimizes AI agents and language models in containerized environments, supporting large-scale parallel benchmarks and RL rollouts. Integrates with third‑party providers for thousands of parallel environments and serves as the official harness for Terminal‑Bench.
AI Coding2025
Adds a lightweight, spec-driven workflow so AI coding assistants agree on requirements before code is produced — creates per-change artifacts (proposal, specs, design, tasks), exposes CLI slash-commands, and integrates with 20+ tools for repeatable AI-driven development.
AI Agent2025
Deploys autonomous AI agents that dynamically attack running apps and return validated proof-of-concept exploits instead of static-analysis noise. Specialized agents cover IDOR, injection, SSRF, XSS, and auth flaws, with HTTP proxy and CI/CD hooks.
AI Agent2025
Deep research agent for complex, long-horizon research and prediction tasks. Pairs a 256K context window with up to 300 tool calls per query for web search, extraction, and code execution. Ships as open 30B and 235B models scoring 82.7% on GAIA.
MCP Server2025
Provides an MCP server exposing 30+ trading tools — real-time prices, technical indicators, Bollinger Band scores, Reddit/news sentiment, and backtesting — designed to integrate with Claude/OpenClaw agents for automated market analysis.
AI Agent2025
Framework for building multi-modal AI agents that watch, listen, and reason over live video, pairing vision models (YOLO, Roboflow, Moondream) with LLMs like Gemini and OpenAI. Agents join calls in ~500ms and keep audio/video latency under 30ms.
AI Agent2025
Write repository automation as natural-language markdown that compiles into deterministic GitHub Actions workflows running AI agents. Agents run read-only by default and write only via sanitized safe-outputs. Works with Copilot, Claude, Codex, or Gemini.
AI Agent2025
Wraps the OpenCode CLI with a plan-first workflow: agents propose a plan you approve before any code is written, and a ContextScout step loads your repo's existing patterns so output matches house style, not generic boilerplate.
AI Agent2025
Defines a predictable repository-level instruction file for coding agents, giving teams one place to document workflow rules instead of each tool inventing its own context format.