Skill UI

Defines a continuous autonomous agent loop pattern with quality gates, evaluations, and recovery actions, orchestrating CI/PR controls, RFC breakdown, and assessment cycles to keep release branches stable.

Manage long-lived cloud agents via observability, secure boundaries, lifecycle controls, and controlled rollouts, ensuring reliable recovery, auditability, and integration with PM2/systemd/CI-CD workflows.

This pattern addresses the fundamental limitation of context window size by treating the filesystem as a persistent, external memory. Agents can write large outputs, plans, or findings to files, offloading them from the active context. Instead of carrying all information, agents use targeted search and retrieval tools to load only the necessary context on demand, drastically reducing token bloat and improving performance over long, complex tasks.

This infrastructure provides a robust solution for running powerful AI agents in remote, sandboxed environments. It overcomes the limitations of local execution by ensuring unlimited concurrency, consistent environments, and collaborative features. Use this to build complex background coding agents, multi-client interfaces (e.g., Slack, Web), or scale agent workloads far beyond local machine constraints.

Sets up a new AgentHub collaboration session by scaffolding the .agenthub directory, generating session IDs, configuring evaluation criteria, and optionally capturing a baseline metric so teams can immediately spawn agents and run structured evaluations.

Comprehensive suite of 42 marketing agent skills spanning content, SEO, CRO, channels, growth, intelligence, and sales, built on a shared brand context plus orchestration router and stdlib-only Python tools for campaign automation.

Production-grade toolkit that lets Claude Code agents scaffold Playwright configs, generate/review/fix end-to-end tests, migrate suites, sync with TestRail, and run BrowserStack reports.

This agent functions as a Senior Product Manager expert, providing comprehensive support across the entire product lifecycle. It utilizes over 30 proven PM frameworks (e.g., RICE, MoSCoW) and includes 12 templates and 32 essential SaaS metrics with exact formulas. It is designed for structured analysis, assisting with strategy definition, market discovery, roadmap planning, and data-driven decision-making.

Coordinates complex features via a humanplane-inspired RFC pipeline, breaking work into DAG-based agents with quality gates, merge queues, and unit orchestration to keep large jobs verifiable.

Analyzes Claude Code auto-memory, promotes stable learnings to CLAUDE.md/.claude/rules, extracts reusable skills, monitors MEMORY.md health, and provides review/promotion/extraction commands to keep project knowledge curated and actionable.

This skill provides a systematic methodology for iteratively improving complex skill documentation. By utilizing a reviewer agent, it identifies and categorizes issues (Critical, Major, Minor) in a structured loop. The process ensures that the skill meets high standards of quality, reliability, and functional completeness before deployment, moving beyond simple manual edits.

A comprehensive, multi-phase security audit tool designed to evaluate AI agent skills before deployment. It detects critical vulnerabilities, including prompt injection, malicious code, excessive permissions, secret exposure, and supply chain risks, ensuring the integrity and safety of your agent environment.