Serve a quantized or unquantized LLM checkpoint as an OpenAI-compatible API endpoint using vLLM, SGLang, or TRT-LLM. Use when user says "deploy model", "serve model", "start vLLM server", "launch SGLang", "TRT-LLM deploy", "AutoDeploy", "benchmark throughput", "serve checkpoint", or needs an inference endpoint from a HuggingFace or ModelOpt-quantized checkpoint. Do NOT use for quantizing models (use ptq) or evaluating accuracy (use evaluation).
Prepare and run a release you can explain and undo — exact artifact identity, environment parity, config and secrets, a rollout strategy matched to blast radius, and a preflight that names the abort condition in advance. Use when planning or performing a deploy, building a release path, or when someone says "ship it" and the steps are not written down anywhere. Not for proving the deployed thing is healthy (release-verification) or for getting back out (rollback), and it never treats a green build or an approved plan as permission to deploy.
Deploy applications to DigitalOcean App Platform via GitHub Actions with proper environment management and secrets handling. Use when setting up CI/CD pipelines, configuring staging/production environments, managing deployment secrets, or creating GitHub Actions workflows.
Import, release, and manage Falcon Fusion workflow definitions in a CID. TRIGGER when user asks to import a workflow, release a workflow version, list existing workflows, check for duplicates, or manage workflow definitions. DO NOT TRIGGER for writing YAML (use authoring), executing workflows, or monitoring (use execution).
GCP deployment patterns for Cloud Run, Cloud Build, Terraform, OAuth. Use when deploying services, setting up CI/CD, or troubleshooting deployment issues.
Hosting, deployment, CI/CD, and going live. Activated when Claude works with deployment configs, Dockerfiles, Vercel/Netlify configs, or deployment-related commands.
How this repo is built and served in production (Cloudflare Pages). Read before changing build output, scripts, Node/pnpm version, redirects, headers, or the `site` URL.
Ship workflow: merge main, run tests, review diff, auto-changelog, bisectable commits, push, and create PR. Also handles first-time CI/CD setup, environment configuration, hosting decisions, and native mobile app store submission. Use this skill when deployment is requested, a release is ready to ship, or the user says "ship it."
Conventions de déploiement — CI/CD GitHub Actions, gh CLI, Docker/docker-compose, serveurs Debian/Ubuntu, Postgres. Charger pour déployer une application, écrire un workflow CI/CD, ou diagnostiquer un déploiement.
Produces a release procedure for the current application, writing tasks/deployment.md. Inspects the repo (CI config, .env.example, migrations, queue/scheduler usage) plus tasks/architecture.md when present. Use whenever the user says "deployment plan", "release checklist", "how do we ship this", "prepare for production", or as the end-of-pipeline SDLC step.
Deploy applications to production environments. Use when configuring deployments, creating deployment scripts, managing release processes, or setting up CI/CD pipelines.
Safe release practices when shipping a change to a real environment — pre-deploy checklist, rollback plan, gradual rollout (feature flags/canary), and post-deploy verification. Use when the user asks to deploy, release, ship to production/staging, or plan how a risky change should go out. Not for local git branch management (see branching) or CI pipeline configuration itself — this is about the judgment calls around a release, not the mechanics of any one platform.
Ship it: platform config for Vercel, Netlify, Fly, Railway, AWS, env secrets, checklists, and rollback. Use when deploying applications to cloud environments.
Plain text files in a repository that tell a coding agent how the project works: commands to run, conventions to follow and things to avoid. CLAUDE.md, AGENTS.md, cursor rules and skills are the common kinds.
CLAUDE.md or AGENTS.md?
CLAUDE.md is read by Claude Code. AGENTS.md is an open format that Codex, Cursor and other agents read. Many projects keep one and point the other at it.
What is a skill?
A folder with a SKILL.md that describes one capability, such as filling PDFs or reviewing code. The agent loads it only when the task calls for it.
Can I search my own team's files too?
Your agents already can, over MCP, limited to the files you're allowed to read. Searching them from this page is coming.
NVIDIASkill
