artifact-guidelines
Guidelines for writing reports, organizing files, and generating code artifacts
Best use case
artifact-guidelines is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Guidelines for writing reports, organizing files, and generating code artifacts
Teams using artifact-guidelines should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/artifact-guidelines/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How artifact-guidelines Compares
| Feature / Agent | artifact-guidelines | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Guidelines for writing reports, organizing files, and generating code artifacts
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Report Writing and File Organization This skill provides best practices for all subagents when generating written artifacts, code files, and figures. ## Writing Guidelines Use GitHub-flavored CommonMark markdown for all text outputs (reports, logs, documentation). Never use plain .txt files. Write concisely: - Short paragraphs with complete sentences - Favor insight over exhaustiveness - Use lists sparingly, only when they genuinely clarify (e.g., model assumptions, validation criteria) - Avoid markdown overuse - minimal headers and bold, no excessive formatting In reports: - Lead with key findings or conclusions - Support claims with evidence (plots, statistics, diagnostics) - Reference files with clear relative paths: "As shown in `figures/washout_curves.png`..." or "`washout_curves.png`" if in same directory - Document what you tried, what worked, and what didn't In logs: - Capture decisions and reasoning, not play-by-play execution - Record why you chose certain paths or skipped alternatives - Note failures and how you addressed them Use scratchpad: - You should create local files to write a first draft, including thinking process - Rewrite it to form final output, then delete the temporary local files you created ## Code Organization One logical unit per file: - One model per .stan file - One analysis per .py script - Descriptive names: `fit_hierarchical_model.py` not `model.py` - Self-contained scripts that run independently **Stan only:** All Bayesian models must use Stan via CmdStanPy. Do not use PyMC, NumPyro, Pyro, or other PPLs. Keep it simple: - No deep nesting of directories unless natural grouping exists - Clean up exploratory scripts after consolidating insights into reports - Every file should have a clear purpose ## Figure Organization Use descriptive filenames: - `group_washout_curves.png` not `fig1.png` - `prior_predictive_check.png` not `ppc.png` One figure per concept or question: - Avoid packing too many subplots (max 2x2 for comparisons) - Save at appropriate resolution (300 DPI for reports, 150 for exploratory) ## File Minimalism Generate fewer, better files: - Consolidate related content - one EDA report, not 10 partial analyses - Combine related visualizations into multi-panel figures when appropriate - Only create files that will be read by users or subsequent agents - Remove intermediate artifacts after they've served their purpose Each file you create should justify its existence. Ask: will this be read? Does it convey unique information?
Related Skills
anthropic-web-artifacts-builder
Suite of tools for creating elaborate, multi-component claude.ai HTML artifacts using modern frontend web technologies (React, Tailwind CSS, shadcn/ui). Use for complex artifacts requiring state management, routing, or shadcn/ui components - not for simple single-file HTML/JSX artifacts.
documentation-guidelines
Write or update backend feature documentation that follows a repo's DOCUMENTATION_GUIDELINES.md (or equivalent) across any project. Use when asked to create/update module docs, API contracts, or backend documentation that must include architecture, endpoints, payloads, Mermaid diagrams, and seeding instructions.
developer-guidelines
Guidelines for the Developer role: strict adherence, no unsolicited refactoring, documentation, security.
artifact-creation
Guide the creation of artifacts (rules, skills, commands, subagents) from patterns. Use templates appropriately, ensure proper structure, add cross-references, and update documentation. Use when creating new artifacts from extracted patterns.
artifact-management
Manage build artifacts, Docker images, and package registries. Configure artifact repositories, versioning, and distribution strategies.
artifact-lifecycle
Unified lifecycle management for all framework artifacts (skills, agents, hooks, workflows, templates, schemas)
web-design-guidelines
Review UI code for Web Interface Guidelines compliance. Use when asked to "review my UI", "check accessibility", "audit design", "review UX", or "check my site against best practices".
web-artifacts-builder
Suite of tools for creating elaborate, multi-component claude.ai HTML artifacts using modern frontend web technologies (React, Tailwind CSS, shadcn/ui). Use for complex artifacts requiring state ma...
vercel-web-design-guidelines
Reviews UI code for compliance with web interface best practices. Use when auditing accessibility, reviewing UI/UX patterns, checking performance, or improving web design quality. Triggers on "check my site", "audit design", "accessibility review", "UX best practices", or UI code review tasks. Covers 100+ rules for accessibility, performance, and user experience.
swift-human-guidelines
Comprehensive Swift 6 and SwiftUI development guidelines for building iOS 26, iOS 18, iPadOS, macOS, watchOS, visionOS, and tvOS applications. Covers Foundation Models API, BGContinuedProcessingTask, Call Translation API, Liquid Glass design system, data-race safety, typed throws, synchronization primitives, SwiftUI/UIKit interoperability, zoom transitions, and document-based apps. Use when building new Apple platform apps, implementing Apple Intelligence features, optimizing performance with Swift 6 concurrency, following Apple Human Interface Guidelines, creating cross-platform applications, or working with iOS 26/18 APIs. Triggers on Swift code, SwiftUI views, Xcode projects, app architecture, background processing, translation features, Foundation Models, synchronization, actors, Sendable types, or modern Apple platform development.
react-guidelines
React coding guidelines and best practices. MUST follow these rules. Use when reviewing or writing React code or tasks.
python-backend-guidelines
FastAPI/Django backend patterns and best practices