Menu
SkillsMP has collected 22 skills from marcodsl/nix. Open a skill to review its source and details.
Showing 22 of 22 collected skills.
Write and review LLM prompts and coding-agent instruction files: tighten routing in name/description, behavior defaults, output format constraints, tool-use phrasing, and verification. Triggers: editing system or developer prompts, agent rules, persona files,…
Generate a spread of distinct software-engineering options with concrete tradeoffs, then narrow to one before design. Triggers: 'brainstorm', 'ideas for', 'what are my options', 'how should I approach', 'explore alternatives', 'wide mode', or any prompt where…
Plan, design, or review software changes with structured tradeoff analysis across architecture, implementation quality, validation, and performance, surfacing 2-3 options per concern with explicit recommendation. Triggers: 'review this code', 'PR review',…
Per-language coding rules and project bootstrap with strict verification (lint, type-check, test) and idiomatic design. Triggers: Python (`uv`, `ruff`, `mypy`/`ty`, `pytest`, `pyproject.toml`, PEP 723, `prek`, dependabot, migration from…
Generate or execute a Conventional Commits 1.0.0 message that matches the actual diff. Triggers: 'write a commit message', 'commit this', 'is this commit message OK', `feat:`/`fix:`/`docs:`/`refactor:` prefixes, scope choice like `(api)`/`(parser)`, `BREAKING…
Write and review D2 diagrams so they compile and read clearly: pick layout engine, shape, container nesting, and styling intentionally. Triggers: `.d2` files, `d2 fmt`, `d2lang.com`, choosing between `dagre`/`elk`/`tala`, `sql_table` ER diagrams,…
Write and review Dockerfiles with multi-stage builds, pinned base images, cache-efficient layers, and container hardening. Triggers: `Dockerfile`, `FROM`, `RUN apt-get`, `COPY`/`ADD`, `ENTRYPOINT`/`CMD`, `.dockerignore`, multi-stage refactor, image shrinking,…
Build distinctive, production-grade frontend interfaces in HTML/CSS/JS or React with a deliberate aesthetic direction (brutalist, editorial, retro-futuristic, refined minimal, etc.) so the output does not look like generic AI UI. Triggers: 'make it…
Plan, build, secure, and operate Google Cloud workloads. Triggers: `gcloud`, AlloyDB, BigQuery / BQML, Cloud SQL, Cloud Run (services, jobs, worker pools), Firebase, GKE, Vertex AI / Agent Platform / Gemini API, Google AIP API conventions, Well-Architected…
Detect fail-open insecure defaults that let apps run insecurely in production, distinguishing them from fail-secure code that crashes on missing config. Triggers: `process.env.X || 'default'`, `getenv(...) or '...'`, `ENV.fetch(..., default:)`, hardcoded…
Draw 4 Tarot cards with real crypto-random entropy to break a tie when prompts are vague or casually delegated, then map the spread to a concrete next step. Triggers: 'let fate decide', 'YOLO', 'whatever', 'up to you', 'idk', 'surprise me', 'dealer's choice',…
Idempotent upsert of Linear issues against current work: mark started or done, create sub-issues for partial work, attach new issues to open epics or parents, and dedupe before creating. Defaults to dry-run. Triggers: 'sync to Linear', 'create a Linear…
Edit prose so it reads as human, not AI-generated: replace filler, hedging, marketing voice, and AI tells (em dashes, monotonous openers, padded three-item lists, restated claims, stacked adjectives) with concrete actors, scoped claims, and evidence.…
Write and review Next.js 16+ App Router code with strict verification and idiomatic design, plus the Vercel-Labs Next.js best-practices reference (auto-applied alongside the guidelines). Triggers: `app/` directory, `'use client'`/`'use server'`/`'use cache'`,…
Investigate the working tree, group changes into atomic commits with Conventional Commits messages, and emit a paste-ready bash script. Inspects only; never runs `git add`/`commit`/`push` itself. Triggers: 'plan commits', 'commit my changes', 'stage and…
Drive a real browser from the shell with `playwright-cli`: snapshot pages, interact via element refs, manage tabs and sessions, capture screenshots and traces. Triggers: `playwright-cli` or `npx playwright-cli` commands, snapshot YAML, `playwright-cli…
Interview a plan one focused question at a time, presenting 2-3 options with tradeoffs, until decisions, dependencies, and risks are explicit. Maintains a running `### Current Status` block. Triggers: 'pressure-test this', 'stress-test the design', 'poke…
Write and review React 19+ code with strict verification and idiomatic design: Server Components by default, `'use client'` at the smallest interactive leaf. Triggers: `.tsx`/`.jsx` components, hooks (`useState`, `useEffect`, `useMemo`, `useCallback`,…
Route WorkOS implementation, debugging, migration, and docs-lookup tasks to the correct reference file before answering. Triggers: WorkOS docs URLs, terms, or dashboard fields (Sign-in endpoint, `initiate_login_uri`, Redirect URI, `WORKOS_*` env vars),…
Write, rewrite, or improve Linear issues following the Linear Method: short plain-language titles, optional descriptions, concrete tasks instead of user stories. Prompts to split scope when the issue is too big for one task. Falls back to writing one markdown…
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or…
Guides the design and structuring of workflow-based Claude Code skills with multi-step phases, decision trees, subagent delegation, and progressive disclosure. Use when creating skills that involve sequential pipelines, routing patterns, safety gates, task…