skills/z-skill-creator/SKILL.md
Expert system for creating Skills for the Antigravity agent environment. Generates complete `.agent/skills/` directories with SKILL.md files, helper scripts, examples, and resources following strict structural and writing standards. Use this skill whenever the user asks to build, generate, or scaffold a skill for Antigravity, mentions `.agent/skills/`, asks for a SKILL.md template, or wants to create agent capabilities using the Antigravity skill format. Also trigger when the user says things like 'build me a skill for X' or 'create an Antigravity skill' or references the Antigravity agent system.
npx skillsauth add asets-gobizit/claude-skills z-antigravity-skill-creatorInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Generate high-quality, predictable, and efficient .agent/skills/ directories for the Antigravity agent environment.
.agent/skills/ directory structureEvery skill must follow this folder hierarchy:
<skill-name>/
├── SKILL.md # Required: Main logic and instructions
├── brand_context.md # Recommended: Brand voice, tone, messaging guidelines
├── scripts/ # Optional: Helper scripts
├── examples/ # Recommended: Reference outputs showing target quality
└── resources/ # Optional: Templates or assets
Every skill that generates user-facing content (websites, copy, emails, social posts, documents) should include a brand_context.md file. This file tells the agent HOW to sound, not just WHAT to build.
A good brand_context.md covers:
When creating a skill, ask the user: "Does this skill generate content for a specific brand? If so, I'll create a brand_context.md to keep outputs on-brand."
The examples/ folder should contain reference outputs that show the target quality and style. These act as "gold standard" samples the agent can match against. Include:
The SKILL.md must start with YAML frontmatter:
testing-code, managing-databases). Max 64 chars. Lowercase, numbers, and hyphens only. Never include "claude" or "anthropic" in the name.Example:
---
name: testing-code
description: "Extracts text from PDFs. Use when the user mentions document processing or PDF files."
---
These principles ensure Claude (or any agent consuming the skill) can follow the instructions efficiently:
[See ADVANCED.md](ADVANCED.md)) — only one level deep./ for paths, never \.For complex tasks, the generated skill should include:
--help if unsure.When asked to create a skill, produce output in this format:
### [Folder Name]
**Path:** `.agent/skills/[skill-name]/`
### SKILL.md
---
name: [gerund-name]
description: [3rd-person description with triggers]
---
# [Skill Title]
## When to use this skill
- [Trigger 1]
- [Trigger 2]
## Workflow
[Checklist or step-by-step guide]
## Instructions
[Specific logic, code snippets, or rules]
## Resources
- [Links to scripts/ or resources/]
Then provide supporting files (scripts, examples) if applicable.
If the skill generates user-facing content, also produce:
brand_context.md — Interview the user for brand details (tone, audience, language rules, CTA style) and capture them in this file.examples/[style-reference].md — At least one reference output showing target quality, structure, and tone.Before delivering a skill, verify:
brand_context.md (or user confirmed not needed)examples/ for content-generating skillsdata-ai
Automated backup skill for PKA + Obsidian + Claude memory. Snapshots pka.db (via SQLite .backup so WAL is handled safely), the Obsidian vault, Claude memory files, agent profiles, and the help-content Excel into a single timestamped zip in Zoho WorkDrive. Daily/weekly/monthly retention rotation built in. USE WHEN Danny says "backup", "run backup", "snapshot pka", or to recover from a snapshot.
testing
Run any question, idea, or decision through a council of 5 AI advisors who independently analyze it, peer-review each other anonymously, and synthesize a final verdict. Based on Karpathy's LLM Council methodology, packaged as the stress-test skill. MANDATORY TRIGGERS: 'stress-test this', 'stress test this', 'pressure-test this', 'pressure test this', 'war room this', 'council this', 'run the council', 'debate this'. STRONG TRIGGERS (use when combined with a real decision or tradeoff): 'should I X or Y', 'which option', 'what would you do', 'is this the right move', 'validate this', 'get multiple perspectives', 'I can't decide', 'I'm torn between'. Do NOT trigger on simple yes/no questions, factual lookups, or casual 'should I' without a meaningful tradeoff (e.g. 'should I use markdown' is not a stress-test question). DO trigger when the user presents a genuine decision with stakes, multiple options, and context that suggests they want it pressure-tested from multiple angles.
content-media
Weekly scan of Obsidian vault for empty/low-content notes. Moves them to zToBeDeleted/ for user review. Use when the user says "tidy obsidian", "clean obsidian", "obsidian cleanup", or on scheduled weekly run.
development
Design-first website builder. Accepts a Claude Design export (HTML or screenshot) as a visual blueprint, collects business inputs, calls Claude API to generate a design-matched index.html, and publishes to GitHub Pages. Falls back to vibe-based generation if no design is provided. Use when the user says "make-website-design", "design website", "build from design", or provides a Claude Design export for website generation.