Bookmarks
Top 10 GitHub Repos
- BuilderIO/agent-native
- A framework for building agent-native applications — not agents that chat alongside apps, but agents that act inside them. Provides primitives for shared actions, SQL-backed state, identity, tools, skills, jobs, observability, and UI surfaces that all work together. Notable for being a full-stack approach to agentic software rather than just another agent SDK.
- DeusData/codebase-memory-mcp
- A high-performance code intelligence MCP server that indexes codebases into a persistent knowledge graph. Indexes an average repo in milliseconds, supports 158 languages, sub-ms queries, and uses 99% fewer tokens than naive approaches. Ships as a single static binary with zero dependencies. Can index the Linux kernel (28M LOC) in milliseconds.
- NVIDIA/SkillSpector
- A security scanner for AI agent skills that detects vulnerabilities, malicious patterns, and security risks before installation. Research cited in the repo shows 26.1% of agent skills contain security issues. Works with skills used by Claude Code, Codex CLI, Gemini CLI, and others.
- aws/agent-toolkit-for-aws
- Official AWS-supported MCP servers, skills, and plugins that help AI coding agents build, deploy, and manage applications on AWS. Works with Claude Code, Codex, Cursor, and other major coding agents. Provides the tooling, knowledge, and guardrails for agentic AWS development.
- kunchenguid/no-mistakes
- A local git proxy that intercepts pushes and runs an AI-driven validation pipeline before forwarding to the real remote. Spins up a disposable worktree, validates the branch, and opens a clean PR automatically — but only after every check passes. Designed to "kill all the slop" and raise clean PRs.
- stablyai/orca
- An agent orchestration desktop and mobile app for running a fleet of parallel coding agents. Run Codex, Claude Code, OpenCode, or Pi side-by-side, each in its own worktree, tracked in one place. Includes a mobile companion for monitoring and steering agents from a phone.
- withastro/flue
- A sandbox agent framework from the Astro team — not another SDK, but a programmable TypeScript harness for building autonomous agents and AI workflows. Uses a skill-based architecture with SKILL.md files and a runtime that supports local and remote execution.
- google-labs-code/design.md
- A format specification for describing a visual identity to coding agents. DESIGN.md gives agents a persistent, structured understanding of a design system using YAML frontmatter for design tokens and markdown for guidelines. Enables agents to maintain consistent visual output across sessions.
- agentgateway/agentgateway
- An open-source proxy built on MCP and A2A protocols that provides drop-in security, observability, and governance for agent-to-LLM, agent-to-tool, and agent-to-agent communication. Works across any framework and environment.
- modem-dev/hunk
- A review-first terminal diff viewer built specifically for agent-authored changesets. Features multi-file review with sidebar navigation, inline AI annotations, split/stack layouts, and watch mode. Built on OpenTUI and Pierre diffs.
Honorable Mentions
- mattpocock/skills
- Real-world agent skills from a prominent TypeScript engineer, straight from his .claude directory.
- Panniantong/Agent-Reach
- One CLI to give agents access to Twitter, Reddit, YouTube, GitHub, and more with zero API fees.
- ogulcancelik/herdr
- Terminal-based agent multiplexer with workspaces, tabs, and panes for managing multiple agents.
- vercel-labs/agent-browser
- Fast Rust-based browser automation CLI purpose-built for AI agents.
- mukul975/Anthropic-Cybersecurity-Skills
- 817 structured cybersecurity skills mapped to 6 frameworks for AI agents.