AI Prompts and Skills
A collection of reusable prompts and skills for AI tools in software development.
AI Prompts and Skills
A collection of reusable prompts and skills for AI tools used in day-to-day software development.
About This Collection
This collection contains curated prompts and skills for AI tools that support day-to-day software development activities. Prompts and skills are organized by use case and designed to be integrated into different assistant workflows.
Quick Start
1. Clone the Repository
git clone https://github.com/liatrio-labs/ai-prompts.git
cd ai-prompts
2. Browse Prompts and Skills
Prompts are organized in the prompts/ directory by category:
development/- Development workflow and coding promptsdocumentation-and-research/- Documentation, content generation, and research prompts
Skills are organized in the skills/ directory by capability:
git-commit-conventional/- Conventional commit message and commit hygiene guidancecreate-pull-request/- PR/MR drafting and optional creation workflowcreate-mermaid-diagrams/- Mermaid diagram creation and validation workflowliatrio-brand-guidelines/- Liatrio visual brand standards and UI compliance guidancemastra-api/- Mastra development API debugging and inspection workflowstilt-dev/- Tilt-based local development environment operationsuv-usage/-uvpackage manager workflows and command guidancebranch-surgery-pr-split/- Branch decomposition strategy for splitting oversized or mixed-concern PRs
3. Use Prompts with AI Tools
Option A: Direct Copy Copy prompts directly from the files into your preferred AI tool (Cursor, Claude Code, Windsurf, etc.) and customize the parameters as needed for your specific use case.
Option B: Install with Slash Command Manager (Recommended) Use the Slash Command Manager to automatically install prompts for your AI coding assistant. The tool will detect your installed AI assistants and let you choose where to install:
# Install development prompts
uvx --from git+https://github.com/liatrio-labs/slash-command-manager slash-man generate \
--github-repo liatrio-labs/ai-prompts \
--github-branch main \
--github-path prompts/development
# Install documentation and research prompts
uvx --from git+https://github.com/liatrio-labs/slash-command-manager slash-man generate \
--github-repo liatrio-labs/ai-prompts \
--github-branch main \
--github-path prompts/documentation-and-research
See the Slash Command Manager documentation for more details.
4. Use Skills with npx skills
Install repository skills using the skills.sh CLI:
# List available skills in this repository without installing
npx skills add liatrio-labs/ai-prompts --list
# Add all skills from this repository to your local setup (non-interactive)
npx -y skills add liatrio-labs/ai-prompts --all
# Install all skills from this repository
npx skills add liatrio-labs/ai-prompts --all
# Install specific skills
npx skills add liatrio-labs/ai-prompts --skill git-commit-conventional --skill create-pull-request
# Install to specific agents
npx skills add liatrio-labs/ai-prompts --skill create-mermaid-diagrams -a cursor -a codex
# Install globally instead of project scope
npx skills add liatrio-labs/ai-prompts --skill git-commit-conventional -g
Compatibility notes:
allowed-toolssupport varies by agent, so keep frontmatter portable across assistants.- Use
DISABLE_TELEMETRY=1(orDO_NOT_TRACK=1) to opt out of telemetry when running the CLI. - If internal skills are introduced later, prefix the command with
INSTALL_INTERNAL_SKILLS=1(for example,INSTALL_INTERNAL_SKILLS=1 npx skills add liatrio-labs/ai-prompts --list) to include them in discovery.
Skill Creation Scripts (Agent-Focused)
This repository includes local helper scripts for AI-agent-driven skill authoring workflows.
These scripts are derived from OpenAI's skills/.system/skill-creator in openai/skills: https://github.com/openai/skills/tree/main/skills/.system/skill-creator
# Validate an existing skill
uv run scripts/quick_validate.py skills/git-commit-conventional
# Initialize a new skill scaffold
uv run scripts/init_skill.py my-skill --path skills --resources scripts,references --examples
# Generate or refresh agents/openai.yaml
uv run scripts/generate_openai_yaml.py skills/my-skill
For a fuller workflow and troubleshooting notes, see docs/development.md.
Context Markers Quick Reference
Context Markers are emoji-based verification techniques used to detect context rot and ensure critical instructions are being processed correctly by AI agents. This approach provides immediate visual feedback that instructions are being followed, addressing the systematic degradation of AI performance as context length increases. The technique involves requiring AI responses to begin with specific emoji markers, giving users a coarse check that may help detect when context instructions have been lost due to context window limitations or compaction inefficiencies.
Skills
| Emoji | Skill File | Description |
|---|---|---|
| 🎯 | skills/git-commit-conventional/SKILL.md | Conventional Commit Skill |
| 🚀 | skills/create-pull-request/SKILL.md | Pull Request / Merge Request Description Skill |
| 🎨 | skills/create-mermaid-diagrams/SKILL.md | Create Mermaid Diagrams Skill |
Only skills with marker-based context-rot checks are listed here. Skills without marker rows in this table do not define required response markers.
Prompts
| Emoji | Prompt File | Description |
|---|---|---|
| 📋 | prompts/development/address-pr-review-feedback.md | Fetch PR Review Feedback |
| 🔍 | prompts/development/review-branch.md | Review Branch Against Base |
| 🔄 | prompts/development/squash-merge.md | Squash Merge to Main |
| 📚 | prompts/documentation-and-research/check-docs.md | Check Documentation |
| 🧪 | prompts/documentation-and-research/improve-prompt-with-research.md | Improve Prompt with Research-Backed Best Practices |
| 📖 | prompts/documentation-and-research/review-documentation.md | Review and Improve Codebase Documentation |
Repository marker: 🤖 in AGENTS.md for AI Agent Instructions and configuration.
Skills Catalog
| Skill | File | Description |
|---|---|---|
git-commit-conventional | skills/git-commit-conventional/SKILL.md | Create conventional commits with commit hygiene and pre-commit-aware workflows. |
create-pull-request | skills/create-pull-request/SKILL.md | Draft reviewer-ready PR/MR descriptions and optionally create PRs via CLI. |
create-mermaid-diagrams | skills/create-mermaid-diagrams/SKILL.md | Create, validate, and repair Mermaid diagrams with a deterministic CLI loop. |
liatrio-brand-guidelines | skills/liatrio-brand-guidelines/SKILL.md | Apply Liatrio brand guidance for UI design and brand compliance reviews. |
mastra-api | skills/mastra-api/SKILL.md | Debug Mastra agents, tools, workflows, and runtime state via API workflows. |
tilt-dev | skills/tilt-dev/SKILL.md | Manage and troubleshoot Tilt-orchestrated local development stacks. |
uv-usage | skills/uv-usage/SKILL.md | Provide concise guidance for uv project and dependency workflows. |
branch-surgery-pr-split | skills/branch-surgery-pr-split/SKILL.md | Split oversized or mixed-concern branches into reviewable PR stacks with parity audits and merge sequencing. |
Prompt-to-Skill Migration Map
The following prompts have been migrated to skills:
prompts/development/commit.md->skills/git-commit-conventional/SKILL.mdprompts/development/create-pull-request.md->skills/create-pull-request/SKILL.mdprompts/documentation-and-research/create-mermaid-diagrams.md->skills/create-mermaid-diagrams/SKILL.md
Documentation
- Contributing Guidelines - How to contribute prompts, skills, and improvements
- Development Guide - Prompt and skill authoring workflow and validation
- Code of Conduct - Community expectations and reporting guidance
Contributing
We welcome contributions! Please see CONTRIBUTING.md for guidelines on:
- Adding new prompts and skills
- Improving existing prompts and skills
- Submitting issues and feedback
- Following our commit conventions
License
Copyright 2025 Liatrio
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.