spectre-ship

name	spectre-ship
description	👻 | Autonomous end-to-end: brain dump -> scope -> TDD -> commit -> rebase -> PR
user-invocable	true

ship

Input Handling

Treat the current command arguments as this workflow's input. When invoked from a slash command, use the forwarded $ARGUMENTS value.

ship: Autonomous End-to-End Delivery

Take a brain dump and autonomously produce a reviewable PR. Zero confirmation gates — scope, implement with TDD, sweep, rebase, and open a PR.

Execution Style: Fully autonomous. No user approval gates. Parse intent, build it, ship the PR.

ARGUMENTS

<ARGUMENTS> $ARGUMENTS </ARGUMENTS>

Step (1/8) - Parse Context

• Action — ParseBrainDump: Extract from ARGUMENTS:
  • INTENT_TYPE: feat or fix (infer from context — new behavior = feat, broken behavior = fix)
  • TARGET_BRANCH: Extract if specified (e.g., "rebase onto develop"), default origin/main
  • SCOPE_SUMMARY: 1-2 sentence distillation of what to build/fix
  • RELEVANT_FILES: Any files, components, or areas mentioned
  • CONTEXT: Remaining context, constraints, preferences
• Action — ValidateInput:
  • If ARGUMENTS empty → ask user for brain dump, then proceed autonomously
  • Else → proceed

Step (2/8) - Ensure Worktree

• Action — DetectBranch: git rev-parse --abbrev-ref HEAD
  • If already in a worktree or on a non-main branch (not main, not master) → use current context, proceed
  • If on main or master → use EnterWorktree to create an isolated worktree
• Action — SetBranchName: Capture current branch name as BRANCH_NAME for artifact paths

Step (3/8) - Quick Scope

• Action — DispatchResearch: Spawn parallel lightweight agents:

  • @finder — Locate files related to RELEVANT_FILES and SCOPE_SUMMARY
  • @analyst — Understand the relevant code area, key interfaces, existing patterns

• Action — WriteScopeDoc: Create docs/tasks/{BRANCH_NAME}/concepts/scope.md:

  # Scope: {SCOPE_SUMMARY}
  
  ## Objective
  {1-2 sentences from brain dump}
  
  ## Type
  {feat or fix}
  
  ## In Scope
  - {bullet list of what will be done}
  
  ## Out of Scope
  - {what this explicitly won't touch}
  
  ## Target Branch
  {TARGET_BRANCH}
  
  ## Key Files
  {from research — relevant files and their roles}
  

  Keep it ~20 lines. This is a lightweight scope, not full spectre-scope.

Step (4/8) - Create Tasks

• Action — CreateTasks: Use TaskCreate to create 3-8 tasks proportional to scope complexity.
  • Each task gets: clear subject (imperative), description (what to do + acceptance criteria), activeForm (present continuous)
  • Order tasks by dependency — foundational work first
  • Tasks are ephemeral and operational — no file artifact needed

Step (5/8) - Execute with TDD

• Action — ExecuteLoop: For each task sequentially:

  1. TaskUpdate → in_progress
  2. Load Skill(spectre-tdd) for TDD methodology
  3. Execute: RED (write failing test) → GREEN (minimal implementation) → REFACTOR (clean up)
  4. Commit with conventional format: {INTENT_TYPE}({scope}): {description}
  5. TaskUpdate → completed

  Rules:

  • One commit per task minimum
  • Conventional commit format always
  • TDD methodology for implementation tasks (skip for config/doc-only tasks)
  • If a task reveals new work, create additional tasks rather than scope-creeping the current one

Step (6/8) - Sweep

Inline sweep — same checklist as spectre-sweep, no subagents:

6.1 Diff Sanity Check

• Review full diff for unintentional changes (whitespace-only edits, merge artifacts)
• Verify no accidentally staged files outside the intended scope
• Confirm no secrets, API keys, credentials, or sensitive data in diff

6.2 Logging Audit

• Remove temporary/debug logging (console.log, print, debug flags)
• Preserve intentional logs: errors, critical warnings, key state transitions
• Verify log levels are appropriate for production context

6.3 Code Hygiene

• Remove commented-out code (it's in git history if needed)
• Resolve or document any TODO/FIXME/HACK introduced in this session
• Remove hardcoded test values that should be config/env

6.4 Opportunistic Dead Code Cleanup

• Orphaned imports with no usage in the file
• Unused variables or functions declared but never referenced
• Debug artifacts (debugger statements, leftover TODO/FIXME from this work)

6.5 Lint (Strict)

• Run the project linter and fix all violations — no skipping, no eslint-disable
• Address structural lint issues by refactoring, not suppressing

6.6 Test

• Run affected tests + full test suite
• Fix any failures caused by the changes
• Do NOT write new tests here — that was done in Step 5

6.7 Commit Sweep Fixes

• If sweep produced changes, commit: chore({scope}): sweep cleanup

Step (7/8) - Rebase

• Action — FetchLatest: git fetch origin
• Action — CreateSafetyRef: git branch backup/ship-$(date +%Y%m%d-%H%M%S)
• Action — Rebase: git rebase {TARGET_BRANCH}
  • If conflicts → resolve automatically, favoring target branch conventions
  • Track resolution decisions for PR summary
• Action — VerifyPostRebase:
  • Run linter — fix violations
  • Run full test suite — fix failures
  • Confirm commit count and no unexpected changes

Step (8/8) - Create PR

• Action — PushBranch: git push -u origin {BRANCH_NAME}

• Action — CreatePR: gh pr create with:

  Title: {INTENT_TYPE}({scope}): {SCOPE_SUMMARY} (under 70 chars)

  Body:

  ## Summary
  {From scope doc — objective and what was done}
  
  ## Changes
  {Bulleted list derived from completed tasks}
  
  ## Test Plan
  {Bulleted checklist — what was tested, what to verify manually}
  
  Shipped autonomously via `spectre-ship`
  

• Action — OutputPRUrl: Display the PR URL as the final deliverable

Next Steps

Use Skill(spectre-guide) skill to render the Next Steps footer.