pick-next-shell
Pick the next shell whose dependencies are satisfied and carry it through planning: expand, refine, self-improve, halt. Use when the user asks to "pick next shell", "next shell", "continue project", "what's next", "next implementation step", or "continue with the plan".
What it does
Pick Next Shell
Pick the next shell from .turbo/shells/ whose dependencies are satisfied, then carry it through the planning pipeline: expand → refine → self-improve → halt.
Task Tracking
At the start, use TaskCreate to create a task for each step:
- Scan shells and pick next
- Run
/expand-shellskill - Run
/refine-planskill - Run
/self-improveskill - Mark plan ready
- Halt with next-step instructions
Step 1: Scan Shells and Pick Next
Check terminal conditions first:
- No shells and no plans — tell the user to run
/turboplanfor a new task and stop - No shells, but plans exist with
status: donefor all — report the project is complete and stop - No shells, but plans exist with a non-
donestatus — suggest running/implement-planto finish the remaining plans and stop
If shells exist in .turbo/shells/, glob .turbo/shells/*.md and read each file's YAML frontmatter. A shell's depends_on entry is satisfied when .turbo/plans/<dep-slug>.md exists with status: done in its frontmatter.
- Candidates — shells whose
depends_onare all satisfied - Blocked — shells with one or more unsatisfied
depends_on
If there are no candidates (everything is blocked), report which shells are blocked and which dependencies they're waiting on, then stop.
If multiple candidates exist, pick the one with the lowest shell number (from the NN- prefix /draft-shells gives each file). If ambiguous, use AskUserQuestion to let the user choose.
State the picked shell path and its dependencies before continuing.
Step 2: Run /expand-shell Skill
Run the /expand-shell skill, passing the shell file path. Capture the resulting plan path for Step 3.
Step 3: Run /refine-plan Skill
Run the /refine-plan skill with the plan path from Step 2. Loops until the plan stabilizes.
Step 4: Run /self-improve Skill
Run the /self-improve skill to compound planning learnings.
Step 5: Mark Plan Ready
Update the plan's YAML frontmatter to status: ready.
Step 6: Halt with Next-Step Instructions
Halt with this message:
Plan ready at
<plan path>.Planning context is likely full, and the plan is comprehensive enough to continue fresh. Run
/clear, then/implement-plan <slug>to implement. After that, run/pick-next-shellagain for the next shell.
Rules
- Do not edit plan files directly. Revisions go through
/refine-plan. - Never modify the spec file.
- Do not attempt to auto-implement. The user drives implementation with
/implement-planin a fresh session. - If a shell file is missing or has invalid frontmatter, halt and report.
Capabilities
Install
Quality
deterministic score 0.59 from registry signals: · indexed on github topic:agent-skills · 280 github stars · SKILL.md body (2,653 chars)