Synced package doc
Docs/prompts-gpt Package/prompts-gpt generate-orchestration

prompts-gpt generate-orchestration

Scaffold a new orchestration file locally. Pipeline mode creates a JSON config and step prompt files. Parallel and eval modes create Markdown prompt templates.

Usage

prompts-gpt generate-orchestration [--title <text>] [--goal <text>] [--mode <mode>] [--overwrite]

Options

FlagDescription
--title <text>Orchestration title (used as filename)
--goal <text>What the orchestration should accomplish
--description <text>Short description
--mode <mode>auto, parallel, pipeline, or eval. Default: parallel.
--overwriteReplace existing files
--cwd <path>Project directory

What It Creates (Pipeline Mode)

.prompts-gpt/orchestrations/
  <name>.json                          # Pipeline config
  pipeline-steps/
    research.md                        # Step 1 prompt
    implement.md                       # Step 2 prompt
    test-and-review.md                 # Step 3 prompt

Examples

# Interactive wizard
prompts-gpt generate-orchestration

# Non-interactive pipeline
prompts-gpt generate-orchestration \
  --title "Full Stack Feature" \
  --goal "Add user auth" \
  --mode pipeline

# Then run it
prompts-gpt orchestrate --mode pipeline \
  --steps .prompts-gpt/orchestrations/full-stack-feature.json

See Also