Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.hellofriday.ai/llms.txt

Use this file to discover all available pages before exploring further.

Running the same prompt twice can produce different results. That’s acceptable for ad-hoc conversation, but not for production workflows that need to meet your organizational standards consistently. Skills are structured instruction sets that give agents the context they need to produce consistent, high-quality output every time. They define your coding standards, review criteria, formatting requirements, and domain-specific knowledge — all in versionable Markdown files.

​
What skills contain

A skill is a Markdown file that captures:
  • Coding standards — language conventions, formatting rules, naming patterns
  • Review criteria — what to look for, what to flag, severity levels
  • Domain knowledge — business rules, architecture decisions, context an agent needs
  • Formatting requirements — output structure, templates, expected formats

​
Creating skills

You can create skills in several ways:
  • Write your own — author Markdown files with your standards and instructions
  • Pull from public repositories — thousands of community-published skills are available
  • Build iteratively with an LLM — have an LLM interview you about what you care about and generate a skill from the conversation

​
Publishing and versioning

Skills are published to your Friday instance and versioned. You can:
  • Browse and manage skills in the Studio
  • Publish skills via the CLI with friday skill publish
  • Upload skill folders through the Studio UI (drag and drop)
  • Track version history and roll back to previous versions

​
Using skills in workflows

Attach skills to agents in your workspace.yml to give them the context they need. Different agents in the same job can use different skills — for example, a code review job might have one agent with SQL expertise skills and another with security review skills.
Skills are designed to evolve. The natural language instructions you give to LLMs are effectively the foundation of AI behavior. Skills let you capture, version, and refine those instructions instead of losing them with every new conversation.