latitude-dev/promptl-skill

Claude Code skill for writing PromptL prompts in the Latitude platform

0 stars0 forksUpdated Jan 22, 2026
npx skills add latitude-dev/promptl-skill

README

PromptL Skill

A Claude Code skill that helps LLMs write PromptL prompts for the Latitude platform.

What is PromptL?

PromptL is a versatile, human-readable language that simplifies defining and managing dynamic prompts for LLMs. It offers:

  • Readable syntax - Easy to write and maintain
  • Dynamic flexibility - Variables, conditionals, and loops
  • Powerful logic - Chains, agents, and tool integrations

Installation

Copy the skill to your Claude Code skills directory:

# Clone the repository
git clone https://github.com/latitude-dev/promptl-skill.git

# Copy to Claude Code skills directory
cp -r promptl-skill ~/.claude/skills/promptl

Or manually create the directory and copy the files:

mkdir -p ~/.claude/skills/promptl
cp SKILL.md AGENTS.md ~/.claude/skills/promptl/

Usage

Once installed, the skill will automatically activate when you're working on tasks involving:

  • Writing PromptL prompts
  • Configuring LLM parameters
  • Using variables, conditionals, or loops
  • Creating multi-step chains
  • Defining tools or agents
  • Working with structured JSON output

You can also explicitly invoke it with /promptl.

What's Included

SKILL.md

Quick reference guide with:

  • Syntax categories overview
  • Quick reference for all features
  • When to apply the skill

AGENTS.md

Comprehensive documentation covering:

  1. Prompt Structure - Config section and message content
  2. Configuration - YAML frontmatter options
  3. Messages - System, user, assistant, and tool tags
  4. Variables - Dynamic content with {{ }}
  5. Conditionals - if/else/endif blocks
  6. Loops - for/endfor iteration
  7. Chains and Steps - Multi-step workflows
  8. Tools - Function calling definitions
  9. Agents - Autonomous AI workflows
  10. Prompt References - Reusable snippets
  11. Content Types - Images, files, and more
  12. Mocking - Testing and development
  13. Structured Output - JSON schema validation
  14. Best Practices - Do's and don'ts

Example

Here's a simple PromptL prompt:

---
provider: OpenAI
model: gpt-4o
temperature: 0.7
---

You are a helpful assistant.

<user>
  Tell me a joke about {{ topic }}.
</user>

And a more advanced agent:

---
provider: OpenAI
model: gpt-4o
type: agent
maxSteps: 40
tools:
  - latitude/search
schema:
  type: object
  properties:
    summary:
      type: string
    key_points:
      type: array
      items:
        type: string
  required:
    - summary
    - key_points
---

Research {{ topic }} and provide a structured summary.

Resources

License

MIT License - see LICENSE for details.

Publisher

latitude-devlatitude-dev

Statistics

Stars0
Forks0
Open Issues0
LicenseMIT License
CreatedJan 22, 2026