
# skill Lifecycle management for building, improving, and maintaining codevoyant-compatible skills. ## Requirements - `npx` — required for scaffolding operations ## Commands ### explore — Research existing skills Search agentskill.sh and known registries for existing skills before building something new. ```bash /skill explore # open-ended research /skill explore "linear integration" # targeted search ``` ### new — Scaffold a new skill Create a new skill dir
Unified task runner dispatcher: auto-detects mise, just, task.dev, or package.json scripts and provides commands to run, list, or detect the project task runner. Triggers on: "/tasks", "run task", "list tasks", "detect task runner", "what tasks", "mise run", "just run", "pnpm run", "package.json scripts".
Terraform authoring patterns — directory structure, backend config, workspace-per-environment, variable management, and mise task wrappers for GCP and AWS. Load when writing .tf files, configuring backends, managing workspaces, or running tf-init/tf-plan/tf-apply tasks.
Specification-driven development. Triggers on: "spec new", "spec go", "spec guide", "spec review", "spec refresh", "spec update", "spec clean", "spec allow", "spec help", and all legacy /spec:* trigger phrases. Unified dispatcher — pass a subcommand as the first argument.
UX design workflows: quick wireframes and approach comparisons, full SvelteKit prototypes, or CSS design token synthesis from a live URL. Triggers on: "ux explore", "ux prototype", "ux style-synthesize", "ux allow", "ux help", "quick wireframe", "wireframe this", "explore approach", "single file prototype", "html mockup", "ux prototype", "synthesize styles", "extract design tokens".
SvelteKit feature-slice architecture patterns. Load when adding routes, building feature libs, designing components, or deciding where code belongs.
TypeScript patterns and fixes for type safety. Load when dealing with any, unknown, type casting, or third-party library types.
Engineering management workflows: plan epics with Linear tasks, approve milestones, review roadmaps, update plans, or manage execution permissions. Triggers on: "em plan", "em approve", "em review", "em update", "em allow", "em help", "plan an epic", "engineering project planning", "push milestones to Linear", "engineering roadmap", "eng plan".
GitHub workflows: watch CI pipelines, manage PR review comments, create and publish draft reviews. Triggers on: "gh ci", "watch github ci", "pull pr comments", "push pr comments", "gh draft", "gh resolve", "gh complete", "github review", "submit pr review".
Docker best practices for containerization, dev containers, and Docker Compose. Load when writing Dockerfiles, docker-compose.yml, or mise Docker tasks. Triggers on: Dockerfile, docker-compose.yml, docker build, docker compose, devcontainer, COMPOSE_BAKE, GCP Artifact Registry, Cloud Run.
End-to-end pipeline orchestration. Chain multiple skill workflows (dev explore, spec new, spec go, dev explore review) into a named flow that runs sequentially. Triggers on: 'flow new', 'flow go', 'create a flow', 'run flow'.
Product management workflows: explore product areas, plan quarterly roadmaps, write PRDs, approve initiatives in Linear, review roadmap quality, or update plans. Triggers on: "pm plan", "pm prd", "pm explore", "pm approve", "pm review", "pm update", "pm allow", "pm help", "product roadmap", "quarterly plan", "write a PRD", "product requirements", "product planning", "explore product area".
Code review workflows: create a draft PR/MR, generate AI-powered inline review comments, address change requests, or complete a draft review. Triggers on: "pr open", "pr new", "pr review", "pr address", "pr complete", "open a PR", "create a draft PR", "code review", "pr mr", "pr this PR", "address pr comments", "fix review comments", "complete draft review", "publish review".
High-performance computing — threading, SIMD, GPU compute with CUDA/SYCL, MPI clustering, Python parallelism, and distributed Ray. Load when working with parallel algorithms, GPU kernels, multi-node clusters, or performance-critical Python/C++ code.
mise.toml authoring, tool management, task patterns, and GCP publish workflows. Load when writing or editing mise.toml, adding tools, defining tasks, or working with mise-tasks/ directories. Triggers on: mise.toml, .mise.toml, mise tasks, mise run, mise install, tool versions, task depends.
Computational geometry — meshes, voxels, point clouds, and signed distance fields. Load when working with 3D formats (STEP, OBJ, STL, glTF), spatial analysis (BVH, raycasting, bounding volumes), or frontend 3D visualization (three.js, react-drei, threlte).
AWS infrastructure patterns with Terraform. Load when writing .tf files targeting AWS, setting up S3/DynamoDB Terraform backend, deploying ECS services, Lambda functions, static sites with CloudFront, Ray clusters, or Firecracker VM fleets.
C++ project patterns: CMake build system, Conan package management, gRPC services, code standards, and release/sanitizer build profiles. Load when writing CMakeLists.txt, conanfile.py, .proto files, or setting up C++ CI.
ML engineering lifecycle: data pipelines with Ray Data, distributed training with Ray Train, model evaluation, TensorBoard observability, MLflow experiment tracking, and model publishing. Triggers on: "mle", "ml engineering", "data pipeline", "distributed training", "model evaluation", "mlflow", "tensorboard", "model publishing".
Find and download SVG icons from svgrepo.com, recolored to the project's brand color. Triggers on: 'find icon', 'icon for', 'skill icon', 'download svg icon', 'icons find', 'icons use'.
GitLab workflows: watch CI pipelines, manage MR review comments, create and publish draft reviews. Triggers on: "glab ci", "watch gitlab ci", "pull mr comments", "push mr comments", "glab draft", "glab resolve", "glab complete", "gitlab review", "submit mr review".
Git version control workflows: create conventional commits or run interactive rebase. Triggers on: "git commit", "git rebase", "git help", "create a commit", "conventional commit", "commit my changes", "interactive rebase", "squash commits".
Developer workflows: plan feature architecture, explore technical approaches, compare repos, generate docs, or approve tasks in Linear. Triggers on: "dev plan", "dev explore", "dev diff", "dev docs", "dev approve", "dev allow", "dev help", "architecture plan", "plan architecture", "technical design", "explore options", "research approaches", "compare repos".
QA workflows: investigate and document bugs, post issues to GitHub/GitLab/Linear, and run browser-agent smoke tests. Triggers on: 'qa debug', 'qa report', 'qa smoke', 'run smoke test', 'report bug', 'investigate issue'.
Python project patterns: uv package/workspace management, MLflow experiment tracking, Ray distributed computing, Nvidia Warp GPU kernels, Pydantic validation, Click CLIs, and service architecture. Load when writing Python with pyproject.toml or uv.lock.
React patterns: Zustand state management, shadcn/ui + Tailwind CSS, React Three Fiber + Drei for 3D, folder structure, data fetching, and TypeScript conventions. Load when working on React projects (*.tsx) without SvelteKit.
Unified skill lifecycle management. Triggers on: "skill explore", "skill new", "skill learn", "skill consolidate", "skill update", "skill critique", "skill feedback", "skill help", "skill create", "skill improve", "skill review", "skill scaffold", "explore skills", "create a skill", "make a skill for", "learn from", "extract skill from", "skill from PR", "skill from URL", "skill from path", "merge skills", "consolidate skills", "combine skills", "update skill", "improve skill", "critique skill", "review skill quality", "report skill bug", "skill issue", "scaffold skill repo", "init skill repo", "new skill repo". Pass a subcommand as the first argument.
LLM engineering: agent frameworks (AI SDK, LangGraph), document and image processing, open-weight model serving (vLLM on AWS/GCP), RAG pipelines, and eval tooling (RAGAS, LangSmith). Triggers on: "llm", "llm engineering", "agent framework", "rag", "retrieval augmented generation", "model serving", "vllm", "langchain", "langgraph", "ragas", "langsmith", "llm eval".
Linear issue management. Create bug reports and issues in Linear projects and teams. Triggers on: 'linear report', 'report to linear', 'create linear issue', 'log bug in linear'.
GCP context skill — gcloud authentication, Artifact Registry image push/pull, Cloud Run deploy, Terraform infra patterns, and common mise.toml GCP env vars. Triggers on: ".tf files", "Terraform", "Dockerfile with gcr.io or pkg.dev", "mise.toml with GCP_ env vars", "gcloud commands", "Cloud Run", "Artifact Registry", "GCP project", "service account impersonation".