naga_control
Naga's own control capabilities. Pause/resume voice and Live2D, switch models/characters, manage sessions/skills, control music playback, start/stop travel, etc. Use when the user requests control of its own functions.
Skills
Find an Agent Skill for your task. Read the guide and copy the install command.
Updated 11-05 00:16
Agents & models
25,654 results
use-agent-browser
Standardize the use of agent-browser (AI browser automation CLI/MCP) to complete page opening, element interaction, screenshot evidence capture, and Web debugging: installation and doctor diagnostics, open/snapshot/@ref workflow, waiting and viewport strategy, Windows startup failure fallback chain (--no-sandbox, CDP takeover, headless...
occ-verify-release
Verify Open Claude Code package integrity, release tests, and automation evidence. Use when asked whether an OCC release is published, tested, or fully…
clarify
Decomposes underspecified requests by classifying gaps (missing info, ambiguity, false premises) and generating targeted clarifying questions. Produces a…
prompt
Design and review prompts Claude-first — system prompts, reusable agent definitions, long or novel dispatch briefs, pipeline stages, and prompts users will run…
anthropic-skill-creator
Design, review, and improve Claude/Codex skills based on Anthropic's "The Complete Guide to Building Skills for Claude". Use when creating a new skill,…
sablier-create-airdrop
This skill should be used when the user asks to create "token airdrops", "Merkle airdrops", "community airdrops", "Sablier airdrop", "vested airdrops",…
codex-computer-use
Run Codex non-interactively to complete desktop GUI tasks with Computer Use. Use when the user asks to delegate a Mac app/browser UI task to a Codex…
model-id-lookup
Skill for querying AI model IDs. Use this skill when the user needs to find or verify an AI model ID. It supports querying the local cache first, and if not found, updating the local data from the online models.dev and then querying again.
ralph
Recursive task execution - works with any agent harness. Triggers on: ralph, run ralph, recursive handoff. Executes tasks one at a time, then creates new…
agent-incident
Triage and manage incident.io incidents, alerts, schedules, escalations, and status pages. Use when the user asks about active incidents, who's on-call, alert…
agent-vercel
Triage Vercel deployments, builds, logs, domains, and spend: what is live in production including rolling releases, build and runtime logs, env var diffs…
ask-dp
DeepWorks skill closed-loop router. You won't remember how every dp-* skill in the company works together, so ask. When the user says "use DeepWorks to do something but I don't know where to start", "run through the device diagnostic evaluation", "modify the knowledge base and then verify", "find out why the agent answered that way", or any task that requires chaining multiple dp-* skills, first use this map to decide the entry point; you can also first run...
dp-ask-teams
Initiate a task to an Agent Teams cloud agent and wait until it completes to get the result, not limited to a specific domain expert. Triggered when the user says "initiate a cloud task", "let a cloud expert do this", "call Agent Teams to run this", "run it in the cloud", or wants to directly invoke the cloud agent-teams capability. When --agent-name is not provided, list all available experts. token…
ai-red-teaming
Run structured AI red team exercises for jailbreak resistance, data exfiltration risk, harmful output controls, and agent tool abuse resilience.
multiplayer
Multiplayer game development principles. Architecture, networking, synchronization.
recsys-pipeline-architect
Designs composable recommendation, ranking, and feed pipelines using the six-stage Source→Hydrator→Filter→Scorer→Selector→SideEffect framework
yao-meta-skill
Create, refactor, evaluate, and package agent skills from workflows, prompts, transcripts, docs, or notes. Use for skill creation, reusable workflow packaging,…
configure-skill-invocation
Make selected global or project skills available only through explicit invocation.
kinde-mcp
Kinde MCP: connect, inspect, and safely operate Kinde's official MCP products. Use when configuring the Kinde Management MCP server in a supported client;…
reverse-claude
The reverse-engineering manual for Claude Code. Encodes all seven runtime systems (permissions, plugins, agents, compaction, self-configuration, query engine…
soia-env-storage-cleanup
For beginners, measure the space usage of SOIA managed configuration, state, cache, and temporary directories, generate a cleanable list, and warn about deletion risks; only after the customer has reviewed the latest list and explicitly authorized it may deletion be performed, followed by verification of the actual space freed. Triggers: "Check SOIA usage", "Measure cache size", "Clean temporary files", "Clean expired state", "Free up disk space"; not for unauthorized full-disk cleanup.
xuefeng-method
Xuefeng-style AI-Native product development methodology. Applicable to: (1) AI-native products with open, non-exhaustible user behavior (AI calendars, AI assistants, AI recommendations, conversational products, etc.), (2) strongly model-dependent scenarios, where AI drives core decisions rather than merely assisting, (3) design and division of labor for multi-specialist Agent architectures, (4) rapid calibration, behavior auditing, and drift detection after launch, (5)…
mastra-bedrock
Use Amazon Bedrock as the model provider for a Mastra agent, and deploy Mastra agents to AWS Bedrock AgentCore. Use when someone wants a Mastra agent to run on…