Loading...
Loading...
Found 2,223 Skills
SOUL 自演进技能 — 在不可变段保护下更新可演进人格/协作策略并支持回滚。
Orchestrate parallel scientist agents for comprehensive analysis with AUTO mode
Interactive session to craft a system prompt for an AI agent powered by Sanity Agent Context MCP.
Terminal-Bench integration for Mux agent benchmarking and failure analysis
Hand off a task to Codex CLI for autonomous execution. Use when a task would benefit from a capable subagent to implement, fix, investigate, or review code. Codex has full codebase access and can make changes.
Trading personality and arena behavior for pump.fun token trading. Governs how the agent trades, announces trades, reacts to other agents, and handles wins and losses in the shared arena group.
Manages organizational guidelines, policies, and best practices as governance variables accessible to all AI agents via SmartContext. Use when working with company rules, brand voice, compliance policies, playbooks, or when any task needs organizational context before proceeding.
Spawn claude, codex, or gemini CLI workers in tmux panes for parallel task execution
Explore-first wave pipeline. Decomposes requirement into exploration angles, runs wave exploration via spawn_agents_on_csv, synthesizes findings into execution tasks with cross-phase context linking (E*→T*), then wave-executes via spawn_agents_on_csv.
This skill should be used when the user requests to "initialize team", "create development team", "team init", "form a team", or "start project team". It collects project information through interactive Q&A and creates an Agent engineering team with professional roles. 8 team types are supported: software development, software testing, reverse engineering, debugging/bug fixing, security research, CTF competition, software and server operation & maintenance, discussion/seminar.
Playwright-based browser automation via Chrome extension + MCP/CLI. Connects to your RUNNING browser (existing logins, cookies, extensions preserved). Use for authenticated flows, stateful web automation, and AI agent browser control without re-logging in.
Detect and neutralize prompt injection attacks in OpenClaw skill content, user inputs, and external data sources. Prevents instruction hijacking and context manipulation.