L0: ln-1000-pipeline-orchestrator (sequential Skill calls, single context)
  +-- Skill("ln-300") — task decomposition (internally manages its own workers)
  +-- Skill("ln-310") — validation (internally launches Codex/Gemini agents)
  +-- Skill("ln-400") — execution (internally dispatches Agent(ln-401/403/404), Skill(ln-402))
  +-- Skill("ln-500") — quality gate (internally runs ln-510/ln-520, verdict, finalization)

Backlog       --> Stage 0 (ln-300) --> Backlog      --> Stage 1 (ln-310) --> Todo
(no tasks)        create tasks         (tasks exist)      validate            |
                                                          | NO-GO             |
                                                          v                   v
                                                       [retry/ask]    Stage 2 (ln-400)
                                                                             |
                                                                             v
                                                                      To Review
                                                                             |
                                                                             v
                                                                      Stage 3 (ln-500)
                                                                       |          |
                                                                      PASS       FAIL
                                                                       |          v
                                                                      Done    To Rework -> Stage 2
                                                               (branch pushed)  (max 2 cycles)

IF .pipeline/state.json exists AND complete == false:
  # Previous run interrupted — resume from saved state
  1. Read .pipeline/state.json -> restore: selected_story_id, story_state,
     quality_cycles, validation_retries, story_results,
     stage_timestamps, git_stats, pipeline_start_time, readiness_scores
  2. Read .pipeline/checkpoint-{selected_story_id}.json -> get last completed stage
  3. Re-read kanban board -> verify selected story still exists
  4. IF worktree_dir exists (.worktrees/story-{selected_story_id}): cd {worktree_dir}
  5. Jump to Phase 4, starting from stage AFTER checkpoint.stage

IF .pipeline/state.json NOT exists OR complete == true:
  # Fresh start — proceed to Phase 1

IF storage_mode == "linear":
  statuses = list_issue_statuses(teamId=team_id)
  status_cache = {status.name: status.id FOR status IN statuses}

  REQUIRED = ["Backlog", "Todo", "In Progress", "To Review", "To Rework", "Done"]
  missing = [s for s in REQUIRED if s not in status_cache]
  IF missing: ABORT "Missing Linear statuses: {missing}. Configure workflow."

pipeline_dir = "$(pwd)/.pipeline"
Write .pipeline/state.json:
  Initialize: complete=false, selected_story_id,
  all counters=0, empty collections,
  business_answers from Phase 2, storage_mode, project_brief, story_briefs,
  status_cache (Linear) or {} (file), pipeline_dir

IF platform == "win32":
  Bash: cp {skill_repo}/ln-1000-pipeline-orchestrator/references/hooks/prevent-sleep.ps1 .pipeline/prevent-sleep.ps1
  Bash: powershell -ExecutionPolicy Bypass -WindowStyle Hidden -File .pipeline/prevent-sleep.ps1 &
  sleep_prevention_pid = $!

branch_check = git branch --show-current
IF branch_check matches feature/* / optimize/* / upgrade/* / modernize/*:
  worktree_dir = CWD
  project_root = CWD
ELSE:
  story_slug = slugify(selected_story.title)
  branch = "feature/{selected_story_id}-{story_slug}"
  worktree_dir = ".worktrees/story-{selected_story_id}"
  project_root = CWD

  changes = git diff HEAD
  IF changes not empty:
    git diff HEAD > .pipeline/carry-changes.patch

  git fetch origin
  git worktree add -b {branch} {worktree_dir} origin/master

  IF .pipeline/carry-changes.patch exists:
    git -C {worktree_dir} apply .pipeline/carry-changes.patch && rm .pipeline/carry-changes.patch
    IF apply fails: WARN user "Patch conflicts -- continuing without uncommitted changes"

  cd {worktree_dir}    # All subsequent Skill calls inherit this CWD

# --- INITIALIZATION ---
id = selected_story.id
quality_cycles = 0          # FAIL->retry counter, limit 2
validation_retries = 0      # NO-GO retry counter, limit 1
story_state = "QUEUED"
story_results = {}          # {stage0: "...", stage1_agents: "...", ...}
stage_timestamps = {}
git_stats = {}
readiness_scores = {}
pipeline_start_time = now()

target_stage = determine_stage(selected_story)    # pipeline_states.md guards

# --- PROGRESS TRACKER (survives compaction) ---
TodoWrite([
  {content: "Stage 0: Task Decomposition (ln-300)", status: "pending", activeForm: "Decomposing tasks"},
  {content: "Stage 1: Validation (ln-310)", status: "pending", activeForm: "Validating story"},
  {content: "Stage 2: Execution (ln-400)", status: "pending", activeForm: "Executing tasks"},
  {content: "Stage 3: Quality Gate (ln-500)", status: "pending", activeForm: "Running quality gate"},
  {content: "Pipeline Report + Cleanup", status: "pending", activeForm: "Generating report"}
])
# Mark each in_progress -> completed as stages execute. Items survive context compaction.

# --- STAGE 0: Task Decomposition ---
IF target_stage <= 0:
  stage_timestamps.stage_0_start = now()
  Skill(skill: "ln-300-task-coordinator", args: "{id}")
  Re-read kanban -> ASSERT tasks exist under Story, count IN 1..8
  IF ASSERT fails: PAUSED, ESCALATE
  stage_timestamps.stage_0_end = now()
  Write stage notes: .pipeline/stage_0_notes_{id}.md (Key Decisions, Artifacts)
  Write checkpoint(stage=0)
  Update .pipeline/state.json

# --- STAGE 1: Validation ---
IF target_stage <= 1:
  stage_timestamps.stage_1_start = now()
  Skill(skill: "ln-310-multi-agent-validator", args: "{id}")
  Re-read kanban -> ASSERT Story status = Todo
  Extract readiness_score from ln-310 output
  IF NO-GO AND validation_retries < 1:
    validation_retries++
    Skill(skill: "ln-310-multi-agent-validator", args: "{id}")    # retry
    Re-read kanban -> ASSERT Story status = Todo
  IF still NOT Todo: PAUSED, ESCALATE
  readiness_scores[id] = readiness_score
  Extract agents_info from .agent-review/review_history.md or ln-310 output
  stage_timestamps.stage_1_end = now()
  Write stage notes: .pipeline/stage_1_notes_{id}.md (Verdict, Agent Review, Key Decisions)
  Write checkpoint(stage=1)
  Update .pipeline/state.json

# --- STAGE 2+3 LOOP (rework cycle) ---
# COMPACTION GUARD: if vars lost after auto-compaction, recover from disk
IF quality_cycles is undefined OR story_state is undefined:
  Read .pipeline/state.json -> restore all vars
  Read .pipeline/checkpoint-{id}.json -> get last completed stage
  Re-read this SKILL.md (full) -> restore Phase 4 flow
  Resume from checkpoint.stage + 1

WHILE quality_cycles < 2:

  # STAGE 2: Execution
  IF target_stage <= 2 OR quality_cycles > 0:
    stage_timestamps.stage_2_start = now()
    Skill(skill: "ln-400-story-executor", args: "{id}")
    Re-read kanban -> ASSERT Story status = To Review AND all tasks = Done
    IF ASSERT fails: PAUSED, ESCALATE, BREAK
    git_stats[id] = parse `git diff --stat origin/master..HEAD`
    stage_timestamps.stage_2_end = now()
    Write stage notes: .pipeline/stage_2_notes_{id}.md (Key Decisions, Git commits)
    Write checkpoint(stage=2)
    Update .pipeline/state.json

  # STAGE 3: Quality Gate (IMPOSSIBLE TO SKIP — next line after Stage 2)
  stage_timestamps.stage_3_start = now()
  Skill(skill: "ln-500-story-quality-gate", args: "{id}")
  Re-read kanban -> check Story status
  Extract quality verdict, score from ln-500 output
  Extract agents_info from .agent-review/review_history.md or ln-500 output
  stage_timestamps.stage_3_end = now()
  Write stage notes: .pipeline/stage_3_notes_{id}.md (Verdict, Score, Agent Review, Branch)
  Write checkpoint(stage=3, verdict, score)
  Update .pipeline/state.json

  IF Story status = Done:
    story_state = "DONE"
    BREAK

  IF Story status = To Rework:
    quality_cycles++
    IF quality_cycles >= 2:
      story_state = "PAUSED"
      ESCALATE: "Quality gate failed {quality_cycles} times. Manual review needed."
      BREAK
    target_stage = 2    # loop back to Stage 2
    CONTINUE

story_state = story_state OR "DONE"    # default if loop exits normally

# 0. Signal pipeline complete
Write .pipeline/state.json: { "complete": true, ... }

# 1. Self-verify against Definition of Done
verification = {
  story_selected:   selected_story_id is set
  story_processed:  story_state IN ("DONE", "PAUSED")
}
IF ANY verification == false: WARN user with details

# 2. Read stage notes
stage_notes = {}
FOR N IN 0..3:
  IF .pipeline/stage_{N}_notes_{id}.md exists:
    stage_notes[N] = read file content
  ELSE:
    stage_notes[N] = "(no notes captured)"

# 3. Extract branch info
branch_name = git branch --show-current
git_stats_final = git diff --stat origin/master..HEAD (if not already captured)

# 4. Finalize pipeline report
durations = {N: stage_timestamps.stage_{N}_end - stage_timestamps.stage_{N}_start
             FOR N IN 0..3 IF both timestamps exist}

Write docs/tasks/reports/pipeline-{date}.md:

  # Pipeline Report -- {date}

  **Story:** {id} -- {title}
  **Branch:** {branch_name}
  **Final State:** {story_state}
  **Duration:** {now() - pipeline_start_time}

  ## Task Planning (ln-300)
  | Tasks | Plan Score | Duration |
  |-------|-----------|----------|
  | {N} created | {score}/4 | {durations[0]} |

  {stage_notes[0]}

  ## Validation (ln-310)
  | Verdict | Readiness | Agent Review | Duration |
  |---------|-----------|-------------|----------|
  | {verdict} | {score}/10 | {agents_info} | {durations[1]} |

  {stage_notes[1]}

  ## Implementation (ln-400)
  | Status | Files | Lines | Duration |
  |--------|-------|-------|----------|
  | {result} | {files_changed} | +{added}/-{deleted} | {durations[2]} |

  {stage_notes[2]}

  ## Quality Gate (ln-500)
  | Verdict | Score | Agent Review | Rework | Duration |
  |---------|-------|-------------|--------|----------|
  | {verdict} | {score}/100 | {agents_info} | {quality_cycles} | {durations[3]} |

  {stage_notes[3]}

  ## Pipeline Metrics
  | Wall-clock | Rework cycles | Validation retries |
  |------------|--------------|-------------------|
  | {total_duration} | {quality_cycles} | {validation_retries} |

# 5. Show pipeline summary to user
Pipeline Complete:
| Story | Branch | Planning | Validation | Implementation | Quality Gate | State |
|-------|--------|----------|------------|----------------|-------------|-------|
| {id} | {branch} | {stage0} | {stage1} | {stage2} | {stage3} | {story_state} |

Report saved: docs/tasks/reports/pipeline-{date}.md

# 6. Worktree cleanup
cd {project_root}
IF story_state == "PAUSED" AND worktree_dir exists AND worktree_dir != project_root:
  git -C {worktree_dir} add -A
  git -C {worktree_dir} commit -m "WIP: {id} pipeline paused" --allow-empty
  git -C {worktree_dir} push -u origin {branch}
  git worktree remove {worktree_dir} --force
  Display: "Partial work saved to branch {branch} (remote). Worktree cleaned."
IF story_state == "DONE" AND worktree_dir exists AND worktree_dir != project_root:
  # ln-500 committed + pushed in Phase 7. Clean worktree only.
  git worktree remove {worktree_dir} --force

# 7. Stop sleep prevention (Windows)
IF sleep_prevention_pid:
  kill $sleep_prevention_pid 2>/dev/null || true

# 8. Remove pipeline state files
Delete .pipeline/ directory

# 9. Report results location to user

## Pipeline Plan for {date}

> **BEFORE EXECUTING -- MANDATORY READ:** Load `{skill_repo_path}/ln-1000-pipeline-orchestrator/SKILL.md` (full file).
> After reading SKILL.md, start from Phase 3 (Pipeline Setup) using the context below.

**Story:** {ID}: {Title}
**Current Status:** {status}
**Target Stage:** {N} ({skill_name})
**Storage Mode:** {file|linear}
**Project Brief:** {name} ({tech})
**Business Answers:** {answers from Phase 2, or "none"}
**Skill Repo Path:** {skill_repo_path}

### Execution Sequence
1. Read full SKILL.md + references (Phase 3 prerequisites)
2. Setup worktree + state.json (Phase 3)
3. Execute stages sequentially via Skill() calls (Phase 4)
4. Generate pipeline report (Phase 5)
5. Cleanup worktree + state files (Phase 5)