Skill UI

Skill Security Auditor scans AI agent skills before installation, flagging command injections, prompt tampering, dependency risks, and filesystem abuses while providing verdicts and remediation guidance for CLIP plugins or Codex skills.

Converts CLAUDE.md system instructions into AI-native structured labels so LLMs follow every rule with fewer tokens, perfect for migrating rules across Claude, Codex, Gemini, or Grok while boosting compliance.

Coordinate multiple AI agents concurrently via dmux’s tmux pane manager, running Claude Code, Codex, OpenCode and other tools in parallel tracks for research, implementation, testing, review, or multi-role coding tasks.

Autonomously orchestrates multi-round research reviews via Codex MCP: review, implement fixes, and re-review until the reviewer declares the work ready or the max rounds expire. Logs progress, maintains compact/recoverable state, and optionally adds human checkpoints for iterative improvement requests.

Brainstorms, validates, and ranks publishable research directions by surveying recent literature, leveraging Codex-powered reasoning, and filtering ideas for novelty, feasibility, and impact before proposing actionable pilots.

Provides a Codex-powered workflow that orchestrates multi-round, high-reasoning GPT research reviews, capturing context, criticisms, follow-ups, experiments, and documentation.