Pick Next Prompt

Pick the next prompt from .turbo/prompts.md and plan its implementation.

Step 1: Plan Style

Run the /plan-style skill to load planning conventions.

Step 2: Read the Prompt Plan

Read .turbo/prompts.md. Parse all prompts extracting:

• Prompt number and title
• Status (pending, in-progress, done)
• Dependencies (Depends on field)
• The prompt text block

Also read .turbo/spec.md (or the spec path referenced in the prompt plan header) for full context.

Step 3: Pick the Next Prompt

Find the first pending prompt whose dependencies are all done.

• If found: proceed to Step 4
• If all prompts are done: report completion to the user — the plan is finished
• If remaining prompts are blocked: report which prompts are blocked and by what

Step 4: Adapt the Prompt

Re-read the spec and compare against the current project state. Adjust the prompt if implementation has diverged:

• File paths changed — update references to match actual structure
• Architecture evolved — adjust the prompt to work with what was actually built
• Scope shifted — add or remove items based on what prior work produced
• New information — incorporate discoveries from previous sessions

Update .turbo/prompts.md with adjustments to the selected prompt. For affected future prompts, update their Context and Depends-on fields if the adaptation changes what they can assume.

Mark the selected prompt in-progress in .turbo/prompts.md.

Step 5: Plan

Using the selected prompt as the requirements, explore the codebase, design the implementation, and write a detailed plan (exact file paths, function signatures, data flow, test cases).

The plan's final step must instruct: "Mark prompt N as done in .turbo/prompts.md."

Rules

• Never modify the spec file — only .turbo/prompts.md is writable (besides the plan file)
• If the prompt plan file does not exist, tell the user to run /create-prompt-plan first
• Keep adaptations minimal — adjust only what diverged, do not rewrite prompts unnecessarily