
Create Meta Prompts
This is for building prompts that talk to other prompts. You give it a task and it generates a Claude prompt optimized for multi-stage workflows, where one prompt's output becomes another's input. It asks you whether you're researching, planning, doing, or refining, then creates structured prompts with XML metadata and creates a SUMMARY.md so you can scan results without reading full outputs. Everything lives in numbered folders under .prompts/ with dependency tracking. The workflow is thoughtful: it detects existing research or plans to chain from, asks clarifying questions based on your purpose, then gives you a decision tree for running the prompt immediately or editing first. Useful when you're orchestrating Claude pipelines instead of writing one-off prompts.
- skills.sh installs
- 176
- GitHub stars
- 2K
- last updated
- Aug 21, 2026
Install
$ npx -y skills add glittercowboy/taches-cc-resources --skill create-meta-prompts --agent claude-codeInstalls into .claude/skills of the current project.
Files
Every execution produces a SUMMARY.md for quick human scanning without reading full outputs.
Each prompt gets its own folder in .prompts/ with its output artifacts, enabling clear provenance and chain detection.
<quick_start>
- Intake: Determine purpose (Do/Plan/Research/Refine), gather requirements
- Chain detection: Check for existing research/plan files to reference
- Generate: Create prompt using purpose-specific patterns
- Save: Create folder in
.prompts/{number}-{topic}-{purpose}/ - Present: Show decision tree for running
- Execute: Run prompt(s) with dependency-aware execution engine
- Summarize: Create SUMMARY.md for human scanning
Sponsors
Details
- Repository
- github.com/glittercowboy/taches-cc-resources
- Categories
- Data Science & ML, Productivity & Planning
- First seen
- Jun 3, 2026