Skill reviews

Draws flat vector illustrations as SVG cards (480x360): slim, long-limbed people acting out one design or coding moment with one hero prop. Identified by no outlines on big shapes,

• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• 11 top-level sections: this looks like several domains in one skill
• The skill tells the model to perform an irreversible action with no human approval

Diagnose and resolve GKE TPU slice creation failures (SliceCreationFailed) caused by partition overlap conflicts where requested TPU partitions or their nodes are already used by a

• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• The skill tells the model to perform an irreversible action with no human approval
• The response is described with custom markup (3 tags): a typed call is more reliable

Run the AgentX agentic serving benchmark with the SemiAnalysis harness. Use for AgentX setup, dataset downloads, concurrency sweeps, or serving latency and throughput comparisons.

• References files that are not bundled: ../deployment/SKILL.md, ../deployment/references/benchmarking.md
• The skill tells the model to perform an irreversible action with no human approval
• 2 referenced file(s) missing: ../deployment/SKILL.md, ../deployment/references/benchmarking.md

Search, summarize, compare, and answer questions from an already configured AutoRAG librarian over local documents and configured datasources. Use when the user asks AutoRAG to sea

• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• The skill tells the model to perform an irreversible action with no human approval
• Does not say what the result is

Create a GitHub tracking issue in Kilo-Org/kilocode that captures a merged Kilo v1 PR's behavior so it can be re-implemented or verified in opencode v2. Use when the user invokes /

• 10 top-level sections: this looks like several domains in one skill
• The skill tells the model to perform an irreversible action with no human approval
• The response is described with custom markup (3 tags): a typed call is more reliable

Adopt and operate LoopX for durable goals, long-running agent work, scoped collaboration, evidence-backed progress, and recovery across sessions. Use when the user asks to set up L

• 12 top-level sections: this looks like several domains in one skill
• The skill tells the model to perform an irreversible action with no human approval
• 7 mutating operations with no state check

Runtime-neutral Reddit account warm-up and brand promotion skill for AI agents. Systematically grow newly registered or existing Reddit accounts into credible identities, then orga

• References files that are not bundled: references/browser-act-rules.md, references/phase1-onboarding.md, references/phase2-preflight.md
• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• 16 top-level sections: this looks like several domains in one skill

Sync delta specs from an OpenSpec change to main specs. Use when the user wants to update main specs with changes from a delta spec, without archiving the change. Also use when the

• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• The skill tells the model to perform an irreversible action with no human approval
• The response is described with custom markup (7 tags): a typed call is more reliable

Bring a published CLI from the public library into the internal library so it's identical to a freshly-generated copy — module path reverted, manuscripts placed alongside, ready fo

• The skill tells the model to perform an irreversible action with no human approval
• The response is described with custom markup (4 tags): a typed call is more reliable
• Does not say what the result is

Draft a reply to one LinkedIn comment from its URL, or sweep a whole thread from just the post URL and draft a reply to every comment worth answering, in one batch. Use for replyin

• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• 13 top-level sections: this looks like several domains in one skill
• The skill tells the model to perform an irreversible action with no human approval

Build and verify the Windows desktop build (NSIS installer + latest.yml) for the canonical OpenMausBot release and its legacy updater mirror. Use when cutting a release, shipping a

• The skill tells the model to perform an irreversible action with no human approval
• The response is described with custom markup (6 tags): a typed call is more reliable
• Does not say what the result is

Design principles and creative direction for building HTML artifacts — pages, reports, plans, landing pages, demos, decks, and small tools. Use when creating or restyling any visua

• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• The skill tells the model to perform an irreversible action with no human approval
• Does not say what the result is

Fine-tune any HuggingFace CV / VLM / LLM model on local NVIDIA GPUs inside an NGC PyTorch container when no dedicated TAO model skill matches. Use when the user wants to fine-tune

• References files that are not bundled: ../../platform/tao-setup-nvidia-gpu-host/SKILL.md, ../../platform/tao-run-on-docker/SKILL.md, references/*.md
• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• The skill tells the model to perform an irreversible action with no human approval

Use when you need to add, remove, edit, list, or change the boot default of an nvpmodel power mode on a Jetson/Tegra (Orin, Thor) target. Triggers: edit power mode, tune frequency

• References files that are not bundled: ../../context/target-platform-contract.md, ../../context/bsp-customization-workflow.md, ../../context/bsp-customization-workflow.md
• 15 top-level sections: this looks like several domains in one skill
• The skill tells the model to perform an irreversible action with no human approval

How to swap the VLM in the VSS Alerts Blueprint — covers RTVI-VLM microservice deployment methods, all three VLM consumers (rtvi-vlm, vlm-as-verifier, vss-agent), and health checks

• 11 top-level sections: this looks like several domains in one skill
• The skill tells the model to perform an irreversible action with no human approval
• The response is described with custom markup (8 tags): a typed call is more reliable

Use when you need to add, remove, edit, list, or change the boot default of an nvfancontrol fan profile on a Jetson/Tegra (Orin, Thor) target. Triggers: edit fan profile, tune fan

• References files that are not bundled: ../../context/target-platform-contract.md, ../../context/bsp-customization-workflow.md, ../../context/bsp-customization-workflow.md
• 15 top-level sections: this looks like several domains in one skill
• The skill tells the model to perform an irreversible action with no human approval

Run a binary industrial-inspection DEFT loop with TAO RT-DETR: measure on frozen KPI/test roles, mine gap-similar real and clean images with SigLIP, accumulate admitted COCO data,

• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• The skill tells the model to perform an irreversible action with no human approval
• 1 test cases, all positive: not one "should refuse" or "should ask first"

Run a TAO training/evaluation/inference container on an NVIDIA Brev GPU instance. Instance provisioning (create/search/stop/delete/login) is delegated to the official brev-cli agen

• The skill tells the model to perform an irreversible action with no human approval
• 1 test cases, all positive: not one "should refuse" or "should ask first"
• Does not say what the result is

Connect the Newton physics engine to an Omniverse simulation or ovstage with ovnewton. Use when importing a populated ovstage into Newton, stepping its physics, synchronizing stage

• The skill tells the model to perform an irreversible action with no human approval
• 7 test cases, all positive: not one "should refuse" or "should ask first"
• No test case covers injection arriving through data

Test system for Megatron-LM. Covers test layout, recipe YAML structure, adding and running unit and functional tests, golden values, marker filters, and CI parity.

• The skill tells the model to perform an irreversible action with no human approval
• The response is described with custom markup (4 tags): a typed call is more reliable
• Does not say what the result is

Scaffold a complete knowledge system. Detects platform, conducts conversation, derives configuration, generates everything. Validates against 15 kernel primitives. Triggers on "/se

• References files that are not bundled: scripts/queries/
• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• The skill tells the model to perform an irreversible action with no human approval

Three.js scene setup: scenes and fog, cameras, the renderer and tone mapping, the animation loop, object hierarchies, transforms, core math and cleanup, on WebGL or with WebGPU and

• 13 top-level sections: this looks like several domains in one skill
• The skill tells the model to perform an irreversible action with no human approval
• The response is described with custom markup (4 tags): a typed call is more reliable

Remote SLURM cluster development via SSH. Use when running jobs, profiling, or developing on a remote SLURM cluster with pyxis/enroot containers. Covers SSH connection management,

• References files that are not bundled: references/<cluster>.md, scripts/build.slurm
• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• 13 top-level sections: this looks like several domains in one skill

Submit and monitor a Slurm job on a local cluster. Supports two modes: (1) Persistent allocation (default) — allocates nodes once via nohup salloc, imports the container once, inst

• Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
• The skill tells the model to perform an irreversible action with no human approval
• The response is described with custom markup (32 tags): a typed call is more reliable

Use when a part has instances above 1 or symmetry, is a left/right pair or a ring of N, or when contract/instance_bbox or instance_count fired. Build repeated, mirrored and radiall

• The skill tells the model to perform an irreversible action with no human approval
• Does not say what the result is
• No condition that starts the skill