realtime-graphics-prompt
Generate a one-shot implementation prompt for an amazing-looking browser game or realtime graphics simulation, synthesizing best practices from researched…
Skills
Find an Agent Skill for your task. Read the guide and copy the install command.
Updated 11-06 00:14
Agents & models
25,710 results
create-skill
Scaffold a new Agent Skill in this marketplace repo with valid SKILL.md, directory layout, and registry entries. Use when adding a skill, creating SKILL.md, or…
north-star-governance
Maintains docs/NORTH_STAR.mdx, wires AGENTS.md as a short agent map, and applies plan hygiene (any format—Superpowers, session plans, etc.) by extracting…
skill-spec-review
Audit SKILL.md and skill directories for Agent Skills spec, Cursor extensions, and npx skills compatibility. Use when reviewing a skill, validating…
handoff
Write a per-session handoff-<session-id>.md for the agent session that resumes this work — in-progress state, exact next steps, and this session's reference…
ai-for-science-ankh-ascend-npu-skill
Ankh protein language model Ascend NPU migration Skill, suitable for migrating Ankh base/large, Ankh3 large/XL, and similar protein models based on HuggingFace Transformers and PyTorch from CUDA/GPU to Huawei Ascend NPU…
external-cannbot-model-model-infer-graph-mode
A skill for adapting Ascend NPU model inference to graph mode based on the PyTorch framework. Adapt models to torch.compile graph mode to accelerate inference performance. Trigger scenarios: npugraph_ex or GE graph mode adaptation, use of torch.compile on Ascend NPU, graph break fixes, aclgraph graph compilation issues. LLM…
external-cannbot-model-model-infer-kvcache
KVCache optimization skill for Ascend NPU model inference based on the PyTorch framework. Analyze and optimize the KVCache implementation in LLM text inference models, including contiguous cache, Paged Attention combined with FA fused operators, and MLA compressed cache. Trigger scenarios include: KVCache management implementation, paged attention, KV...
external-cannbot-model-model-infer-multi-stream
Ascend NPU model inference multi-stream whole-network optimization skill based on the PyTorch framework. Used to analyze and implement multi-stream optimization, dual-stream, stream overlap, core control, and TorchAir multi-stream transformation for models. First perform whole-network module DAG and inter-module parallelism analysis, then perform operator-level analysis, development debugging, and acceptance for each module. Trigger scenarios include: multi-stream optimization, dual-stream, stream…
external-cannbot-model-model-infer-parallel-analysis
A skill for analyzing parallel strategies for Ascend NPU model inference based on the PyTorch framework. It analyzes model architecture parameters and Ascend hardware specifications, and recommends the optimal TP/EP/DP parallel configuration (parallel_config). Trigger scenarios: a new model needs its parallel strategy determined, an existing configuration needs optimization, or reassessment is needed after deployment card count or hardware changes. The output is a structured parallel_config...
external-cannbot-model-model-infer-prefetch
Ascend NPU model inference weight prefetch optimization skill based on the PyTorch framework. Add the torch_npu.npu_prefetch feature to models to optimize inference performance. Trigger scenarios: profiling shows that MatMul/QBMM/GMM operators have memory-bound hotspots, need to add weight prefetch for the model, migrate the prefetch pattern to a new model.
agents
This repo's self-maintained agent definition directory index (complexity-triage / doc-reviewer / doc-updater / tdd-guide), referenced by skills such as design and update when invoking general-purpose subagents to cite definitions and checklists; it is not itself directly callable...
plan-rename
Background hook: auto-rename Claude Code sessions from Plan Mode H1 title, with compaction resilience via sidecar + Stop guard
agent-smith
Orchestrate several coding tasks in parallel from one Herdr pane, with any coding agent (Claude Code, Codex, Antigravity, Grok, OpenCode, Pi, Copilot, Cursor…
find-skills
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express…
kanban-video-orchestrator
Plan, set up, and monitor a multi-agent video production pipeline backed by Hermes Kanban. Use when the user wants to make ANY video — narrative film,…
self-improving-agent
A universal self-improving agent that learns from ALL skill experiences. Uses multi-memory architecture (semantic + episodic + working) to continuously evolve…
tauri-agent-control
Observe and control a running Tauri app via curl over HTTP. Use when asked to interact with the app UI, take screenshots, click buttons, fill inputs, inspect…
agent-memory
Use this skill when the user asks to save, remember, recall, or organize memories. Triggers on phrases like 'remember this', 'save this', 'note this', 'what…
agent-configuration
AI agent configuration policy and security guide. Project description file writing, Hooks/Skills/Plugins setup, security policy, team shared workflow…
agent-email-cli
Operate the agent-email CLI to create disposable inboxes, poll for new mail, retrieve full message details, and manage local mailbox profiles. Use when the…
agent-evaluation
Design and implement comprehensive evaluation systems for AI agents. Use when building evals for coding agents, conversational agents, research agents, or…
agent-tools
Run 150+ AI apps via inference.sh CLI - image generation, video creation, LLMs, search, 3D, Twitter automation. Models: FLUX, Veo, Gemini, Grok, Claude,…
agentic-workflow
Practical AI agent workflows and productivity techniques. Provides optimized patterns for daily development tasks such as commands, shortcuts, Git integration,…