aif-skill-generator

Generates Agent Skills packages with SKILL.md, references, scripts, and security scanning.

31|2|Updated Mar 21, 2026
One-click install
npx skills add https://github.com/letuhao/lore-weave --skill aif-skill-generator-letuhao
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: aif-skill-generator
Source: https://github.com/letuhao/lore-weave/tree/main/.cursor/skills/aif-skill-generator
Command: npx skills add https://github.com/letuhao/lore-weave --skill aif-skill-generator-letuhao

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Creating well-structured Agent Skills that follow the Agent Skills specification is error-prone: frontmatter validation, naming rules, directory layout, and prompt-injection risks are easy to get wrong. This Skill automates the full lifecycle of skill creation, validation, and security scanning. ## Core Features & Use Cases - Skill Generation: Interactively creates complete skill packages (SKILL.md, references, scripts, templates) that conform to the Agent Skills open standard. - Learn Mode: Fetches and studies documentation URLs, then synthesizes the knowledge into a new skill with source attribution. - Two-Level Security Scanning: Combines a regex/static-analysis Python scanner with LLM semantic review to detect prompt injection, data exfiltration, and destructive commands before installing external skills. - Validation Mode: Checks structure, frontmatter YAML, naming conventions, and security in a single combined report. - Use Case: You found a community skill on skills.sh and want to install it. Run the scan command to verify it contains no malicious instructions, then install it only if both security levels pass. ## Quick Start Ask the agent to generate a new skill named 'api-conventions' that documents your team's REST API design guidelines.

Frequently Asked Questions about aif-skill-generator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create a new Agent Skill with SKILL.md?

Invoke the skill with a name to start an interactive workflow that asks about the problem, target user, and tools needed. It then generates a complete package with valid YAML frontmatter, a body under 500 lines, and optional references, scripts, and templates directories.

How do I scan an external skill for prompt injection before installing?

Use the scan command with the skill's path. It runs a two-level check: a Python regex scanner detecting known attack patterns like instruction overrides and data exfiltration, plus a semantic review of all files. Skills with critical findings are blocked and can be removed with the cleanup helper.

Can I generate a skill from documentation URLs?

Yes, Learn Mode activates when you pass one or more URLs. It fetches each page, enriches the material with web searches, synthesizes a knowledge base, asks clarifying questions, and generates a skill with source attribution in the references.

What are the naming rules for Agent Skills?

Skill names must be lowercase with hyphens only, maximum 64 characters, no consecutive hyphens, and must match the directory name. Names with uppercase letters, underscores, or leading hyphens fail validation.

Why does my skill fail validation with argument-hint errors?

Unquoted square brackets in argument-hint break YAML parsing because brackets denote arrays. Wrap the value in quotes, for example argument-hint: "[--flag] <description>", to avoid parse errors and agent TUI crashes.

What happens when a skill is blocked by the security scanner?

Blocked skills are not installed. The cleanup-blocked-skill.py helper removes the installed directory after strict path safety validation and patches skills-lock.json to prevent the blocked skill from being resurrected by later installs.