Total 50,674 skills, AI & Machine Learning has 8493 skills
Showing 12 of 8493 skills
Quality gate via second model. Spawn a different AI model to review work before committing. Includes refusal routing: if one model refuses, silently switch to the next.
Conversationally guides the user to a CLAUDE.md file. Starts by inferring what it can from the repo, asks ~8 core questions, then offers optional depth across business, voice, and working-style topics. Use when the user wants a CLAUDE.md but would rather have a conversation than fill out a form.
For CLI agents WITHOUT subagent support (e.g., Codex CLI). Search previous code agent sessions for specific work, decisions, or code patterns.
Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman. Also auto-triggers when token efficiency is requested. Integrated into Cavekit: enabled by default for build, inspect, and subagent phases via caveman_mode config. See scripts/bp-config.sh for caveman_mode and caveman_phases.
One AI integration. Manage Organizations, Users. Use when the user wants to interact with One AI data.
Use when the user needs self-hosted or local Chroma for semantic search, including `ChromaClient`, `HttpClient`, or Python `EphemeralClient`, local persistence, Docker or `chroma run`, or OSS Chroma without Chroma Cloud features.
AssemblyAI integration. Manage Transcripts, Speakers, Jobs. Use when the user wants to interact with AssemblyAI data.
A method for iteratively improving text instructions for agents (skills / slash commands / task prompts / CLAUDE.md sections / code generation prompts) by having unbiased executors run them, then evaluating from both perspectives (executor self-report + instruction-side metrics). Repeat until improvement plateaus. Use immediately after creating or significantly revising a prompt or skill, or when you suspect the reason an agent isn't behaving as expected is due to ambiguity in the instructions.
Interact with KWeaver Knowledge Network and Decision Agent — build knowledge networks, query Schema/instances, semantic search, execute Action, Agent CRUD and conversation, Trace data analysis. Interact with Dataflow document processes — list processes, trigger runs, query run history, view step logs. Interact with Skill management module — register Skill, search in market, progressive reading, download and installation. Interact with Toolbox / Tool — create toolbox, upload OpenAPI tools, publish, start and stop. Interact with Vega observability platform — query Catalog/resources/connector types, health inspection. This skill is automatically activated when users mention intents such as "knowledge network", "knowledge graph", "query object type", "execute Action", "what Agents are there", "create Agent", "converse with Agent", "list all Agent templates", "list Agents I created", "list Agents in private space", "dataflow", "data flow", "process orchestration", "process run records", "process logs", "trigger dataflow", "view dataflow run history", "Skill", "skill package", "register Skill", "install Skill", "read SKILL.md", "toolbox", "toolbox", "upload tool", "register tool", "OpenAPI tool", "enable tool", "publish toolbox", "data source", "data view", "atomic view", "Catalog", "Vega", "health check", "inspection", "trace", "evidence chain", "data flow tracking", "data source", "how data is obtained", etc.
Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.
Evaluates LLMs across 100+ benchmarks from 18+ harnesses (MMLU, HumanEval, GSM8K, safety, VLM) with multi-backend execution. Use when needing scalable evaluation on local Docker, Slurm HPC, or cloud platforms. NVIDIA's enterprise-grade platform with container-first architecture for reproducible benchmarking.
cuTile Python DSL kernel implementation patterns, CtKernel runtime wrapper, suitability gate, and cuTile-specific pitfalls. Use when: (1) creating or modifying a cuTile Python DSL kernel version, (2) implementing an optimization that still fits within cuTile's exposed control surface, (3) deciding whether cuTile is still the right DSL, (4) reviewing cuTile-specific runtime patterns. Always also load /design-kernel for shared naming, versioning, and workflow.