Vexryn Scan — AI skill for Claude Code
See what your AI coding agent actually loads — and what it can do.
How to install Vexryn Scan
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open yl0n0ps/vexryn-scan and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Vexryn Scan does
See what your AI coding agent actually loads — and what it can do. Any repo, any stack, no config. Reads Claude Code, Cursor, Copilot, Codex, Gemini, Windsurf, Cline, Roo, Continue, Zed, Kiro, OpenCode, Goose. 100% local.
Alternatives in AI
- Memory Forge Rs — Stop resetting satisfying AI chats — edit the memory instead 474 ★
- Waldzell MCP — Waldzell AI's monorepo of MCP servers 202 ★
- AI Toolkit — Professional-grade AI coding toolkit: 94 skills, 44 agents, multi-platform (Claude, Cursor, Windsurf, Copilot 167 ★
README
vexryn
**See what your AI coding agent actually loads — and what it can do.** One command, any repo, no config. Read-only, 100% local, nothing is sent.
[](https://www.npmjs.com/package/vexryn) [](LICENSE)  
npx vexryn scan

Your agent already loaded instruction files, skills and MCP servers before you typed a word — and any config change can quietly hand it new powers: a server holding a token, a tool that can send messages or run shell commands, a package its publisher abandoned. Those configs are scattered across formats and code review has no opinion on them. Vexryn reads them — Claude Code, Cursor, VS Code, GitHub Copilot, Codex, Gemini, Windsurf, Cline, Roo, Continue, Zed, Kiro, OpenCode, Goose — and tells you, in plain words, what each agent loads and what its tools can do. It reads configs, never your code, and never runs a server.
Why it's different
- It reads every agent, in its own format (JSON, TOML, YAML) — not just one.
- It says what tools can do, from a public catalogue of measured servers: send messages, read files, run commands, delete records — and flags the dangerous combinations (read the web + read your files + send them out).
- It reviews a pull request.
vexryn diffturns an unreadable config diff into one comment: new serverslackreceivesSLACK_BOT_TOKEN, unpinned · Claude Code may rungit pushwithout asking · +2,300 tokens every session. - Nothing is sent, nothing is executed, no numbers are invented. An unmeasured server says so. Safe for CI and untruste
Related Skills
Engram
The context spine that 10x's every AI coding session. Live in 8 IDEs (Claude Code, Cursor, Cline, Continue, Ai
Ectype
Read, filter, redact, export and convert the on-disk sessions of 13 AI coding agents (Claude Code, Codex CLI,
Gald3r
AI dev framework: 22 agents, 100 core skills + 142 addon skill packs, 149 commands. Runs on 21 AI tools: Curso
Scrollback
Recall past AI conversations across coding agents — Claude Code, Codex, Devin, OpenCode, Qwen, Gemini, Kimi, F
Agent Provider Configs
Copy-paste configs to point Claude Code, OpenCode, Codex CLI, Cline, Kilo, Aider, Hermes, OpenClaw, Pi, Zed, C
Ctxbox
Local-first AI context manager: scan, search, edit, copy and inject conversations across Claude Code, Codex, C
Related Agents
Consult Agent
Execute cross-tool AI consultations via Task spawning. Use when agents or workflows need a second opinion from
Developer Overview
code-container (container) creates isolated Docker environments for AI coding harnesses (Claude Code, OpenCode
10x Tool Calls
Cursor and Windsurf meter usage by requests and tool calls rather than tokens, which means a finished or stall