execute-phase

name	execute-phase
description	使用波次并行执行阶段中的所有计划。编排器将计划执行委托给子代理，管理波次和检查点。

使用波次并行执行阶段中的所有计划。编排器通过将计划执行委托给子代理保持精简。

<core_principle> The orchestrator's job is coordination, not execution. Each subagent loads the full execute-plan context itself. Orchestrator discovers plans, analyzes dependencies, groups into waves, spawns agents, handles checkpoints, collects results. </core_principle>

<required_reading> Read STATE.md before any operation to load project context. Read config.json for planning behavior settings. </required_reading>

Read model profile for agent spawning:

MODEL_PROFILE=$(cat .gsd/config.json 2>/dev/null | grep -o '"model_profile"[[:space:]]*:[[:space:]]*"[^"]*"' | grep -o '"[^"]*"$' | tr -d '"' || echo "balanced")

Default to "balanced" if not set.

Model lookup table:

Agent	quality	balanced	budget
gsd-executor	opus	sonnet	sonnet
gsd-verifier	sonnet	sonnet	haiku
general-purpose	—	—	—

Store resolved models for use in Task calls below.

Before any operation, read project state:

cat .gsd/STATE.md 2>/dev/null

If file exists: Parse and internalize:

• Current position (phase, plan, status)
• Accumulated decisions (constraints on this execution)
• Blockers/concerns (things to watch for)

If file missing but .gsd/ exists:

STATE.md missing but planning artifacts exist.
Options:
1. Reconstruct from existing artifacts
2. Continue without project state (may lose accumulated context)

If .gsd/ doesn't exist: Error - project not initialized.

Load planning config:

# Check if planning docs should be committed (default: true)
COMMIT_PLANNING_DOCS=$(cat .gsd/config.json 2>/dev/null | grep -o '"commit_docs"[[:space:]]*:[[:space:]]*[^,}]*' | grep -o 'true\|false' || echo "true")
# Auto-detect gitignored (overrides config)
git check-ignore -q .gsd 2>/dev/null && COMMIT_PLANNING_DOCS=false

Store COMMIT_PLANNING_DOCS for use in git operations.

Load git branching config:

# Get branching strategy (default: none)
BRANCHING_STRATEGY=$(cat .gsd/config.json 2>/dev/null | grep -o '"branching_strategy"[[:space:]]*:[[:space:]]*"[^"]*"' | sed 's/.*:.*"\([^"]*\)"/\1/' || echo "none")

# Get templates
PHASE_BRANCH_TEMPLATE=$(cat .gsd/config.json 2>/dev/null | grep -o '"phase_branch_template"[[:space:]]*:[[:space:]]*"[^"]*"' | sed 's/.*:.*"\([^"]*\)"/\1/' || echo "gsd/phase-{phase}-{slug}")
MILESTONE_BRANCH_TEMPLATE=$(cat .gsd/config.json 2>/dev/null | grep -o '"milestone_branch_template"[[:space:]]*:[[:space:]]*"[^"]*"' | sed 's/.*:.*"\([^"]*\)"/\1/' || echo "gsd/{milestone}-{slug}")

Store BRANCHING_STRATEGY and templates for use in branch creation step.

Create or switch to appropriate branch based on branching strategy.

Skip if strategy is "none":

if [ "$BRANCHING_STRATEGY" = "none" ]; then
  # No branching, continue on current branch
  exit 0
fi

For "phase" strategy — create phase branch:

if [ "$BRANCHING_STRATEGY" = "phase" ]; then
  # Get phase name from directory (e.g., "03-authentication" → "authentication")
  PHASE_NAME=$(basename "$PHASE_DIR" | sed 's/^[0-9]*-//')

  # Create slug from phase name
  PHASE_SLUG=$(echo "$PHASE_NAME" | tr '[:upper:]' '[:lower:]' | sed 's/[^a-z0-9]/-/g' | sed 's/--*/-/g' | sed 's/^-//;s/-$//')

  # Apply template
  BRANCH_NAME=$(echo "$PHASE_BRANCH_TEMPLATE" | sed "s/{phase}/$PADDED_PHASE/g" | sed "s/{slug}/$PHASE_SLUG/g")

  # Create or switch to branch
  git checkout -b "$BRANCH_NAME" 2>/dev/null || git checkout "$BRANCH_NAME"

  echo "Branch: $BRANCH_NAME (phase branching)"
fi

For "milestone" strategy — create/switch to milestone branch:

if [ "$BRANCHING_STRATEGY" = "milestone" ]; then
  # Get current milestone info from ROADMAP.md
  MILESTONE_VERSION=$(grep -oE 'v[0-9]+\.[0-9]+' .gsd/ROADMAP.md | head -1 || echo "v1.0")
  MILESTONE_NAME=$(grep -A1 "## .*$MILESTONE_VERSION" .gsd/ROADMAP.md | tail -1 | sed 's/.*- //' | cut -d'(' -f1 | tr -d ' ' || echo "milestone")

  # Create slug
  MILESTONE_SLUG=$(echo "$MILESTONE_NAME" | tr '[:upper:]' '[:lower:]' | sed 's/[^a-z0-9]/-/g' | sed 's/--*/-/g' | sed 's/^-//;s/-$//')

  # Apply template
  BRANCH_NAME=$(echo "$MILESTONE_BRANCH_TEMPLATE" | sed "s/{milestone}/$MILESTONE_VERSION/g" | sed "s/{slug}/$MILESTONE_SLUG/g")

  # Create or switch to branch (same branch for all phases in milestone)
  git checkout -b "$BRANCH_NAME" 2>/dev/null || git checkout "$BRANCH_NAME"

  echo "Branch: $BRANCH_NAME (milestone branching)"
fi

Report branch status:

Branching: {strategy} → {branch_name}

Note: All subsequent plan commits go to this branch. User handles merging based on their workflow.

Confirm phase exists and has plans:

# Match both zero-padded (05-*) and unpadded (5-*) folders
PADDED_PHASE=$(printf "%02d" ${PHASE_ARG} 2>/dev/null || echo "${PHASE_ARG}")
PHASE_DIR=$(ls -d .gsd/phases/${PADDED_PHASE}-* .gsd/phases/${PHASE_ARG}-* 2>/dev/null | head -1)
if [ -z "$PHASE_DIR" ]; then
  echo "ERROR: No phase directory matching '${PHASE_ARG}'"
  exit 1
fi

PLAN_COUNT=$(ls -1 "$PHASE_DIR"/*-PLAN.md 2>/dev/null | wc -l | tr -d ' ')
if [ "$PLAN_COUNT" -eq 0 ]; then
  echo "ERROR: No plans found in $PHASE_DIR"
  exit 1
fi

Report: "Found {N} plans in {phase_dir}"

List all plans and extract metadata:

# Get all plans
ls -1 "$PHASE_DIR"/*-PLAN.md 2>/dev/null | sort

# Get completed plans (have SUMMARY.md)
ls -1 "$PHASE_DIR"/*-SUMMARY.md 2>/dev/null | sort

For each plan, read frontmatter to extract:

• wave: N - Execution wave (pre-computed)
• autonomous: true/false - Whether plan has checkpoints
• gap_closure: true/false - Whether plan closes gaps from verification/UAT

Build plan inventory:

• Plan path
• Plan ID (e.g., "03-01")
• Wave number
• Autonomous flag
• Gap closure flag
• Completion status (SUMMARY exists = complete)

Filtering:

• Skip completed plans (have SUMMARY.md)
• If --gaps-only flag: also skip plans where gap_closure is not true

If all plans filtered out, report "No matching incomplete plans" and exit.

Read `wave` from each plan's frontmatter and group by wave number:

# For each plan, extract wave from frontmatter
for plan in $PHASE_DIR/*-PLAN.md; do
  wave=$(grep "^wave:" "$plan" | cut -d: -f2 | tr -d ' ')
  autonomous=$(grep "^autonomous:" "$plan" | cut -d: -f2 | tr -d ' ')
  echo "$plan:$wave:$autonomous"
done

Group plans:

waves = {
  1: [plan-01, plan-02],
  2: [plan-03, plan-04],
  3: [plan-05]
}

No dependency analysis needed. Wave numbers are pre-computed during /plan-phase.md.

Report wave structure with context:

## Execution Plan

**Phase {X}: {Name}** — {total_plans} plans across {wave_count} waves

| Wave | Plans | What it builds |
|------|-------|----------------|
| 1 | 01-01, 01-02 | {from plan objectives} |
| 2 | 01-03 | {from plan objectives} |
| 3 | 01-04 [checkpoint] | {from plan objectives} |

The "What it builds" column comes from skimming plan names/objectives. Keep it brief (3-8 words).

Execute each wave in sequence. Autonomous plans within a wave run in parallel.

For each wave:

1. Describe what's being built (BEFORE spawning):

   Read each plan's <objective> section. Extract what's being built and why it matters.

   Output:

   ---
   
   ## Wave {N}
   
   **{Plan ID}: {Plan Name}**
   {2-3 sentences: what this builds, key technical approach, why it matters in context}
   
   **{Plan ID}: {Plan Name}** (if parallel)
   {same format}
   
   Spawning {count} agent(s)...
   
   ---
   

   Examples:

   • Bad: "Executing terrain generation plan"
   • Good: "Procedural terrain generator using Perlin noise — creates height maps, biome zones, and collision meshes. Required before vehicle physics can interact with ground."

2. Read files and spawn all autonomous agents in wave simultaneously:

   Before spawning, read file contents. The @ syntax does not work across Task() boundaries - content must be inlined.

   # Read each plan in the wave
   PLAN_CONTENT=$(cat "{plan_path}")
   STATE_CONTENT=$(cat .gsd/STATE.md)
   CONFIG_CONTENT=$(cat .gsd/config.json 2>/dev/null)
   

   Use Task tool with multiple parallel calls. Each agent gets prompt with inlined content:

   <objective>
   Execute plan {plan_number} of phase {phase_number}-{phase_name}.
   
   Commit each task atomically. Create SUMMARY.md. Update STATE.md.
   </objective>
   
   <execution_context>
   ../execute-plan/SKILL.md
   @.gsd/templates/summary.md
   ../../instructions/checkpoints.instructions.md
   ../../instructions/tdd.instructions.md
   </execution_context>
   
   <context>
   Plan:
   {plan_content}
   
   Project state:
   {state_content}
   
   Config (if exists):
   {config_content}
   </context>
   
   <success_criteria>
   - [ ] All tasks executed
   - [ ] Each task committed individually
   - [ ] SUMMARY.md created in plan directory
   - [ ] STATE.md updated with position and decisions
   </success_criteria>
   

3. Wait for all agents in wave to complete:

   Task tool blocks until each agent finishes. All parallel agents return together.

4. Report completion and what was built:

   For each completed agent:

   • Verify SUMMARY.md exists at expected path
   • Read SUMMARY.md to extract what was built
   • Note any issues or deviations

   Output:

   ---
   
   ## Wave {N} Complete
   
   **{Plan ID}: {Plan Name}**
   {What was built — from SUMMARY.md deliverables}
   {Notable deviations or discoveries, if any}
   
   **{Plan ID}: {Plan Name}** (if parallel)
   {same format}
   
   {If more waves: brief note on what this enables for next wave}
   
   ---
   

   Examples:

   • Bad: "Wave 2 complete. Proceeding to Wave 3."
   • Good: "Terrain system complete — 3 biome types, height-based texturing, physics collision meshes. Vehicle physics (Wave 3) can now reference ground surfaces."

5. Handle failures:

   If any agent in wave fails:

   • Report which plan failed and why
   • Ask user: "Continue with remaining waves?" or "Stop execution?"
   • If continue: proceed to next wave (dependent plans may also fail)
   • If stop: exit with partial completion report

6. Execute checkpoint plans between waves:

   See <checkpoint_handling> for details.

7. Proceed to next wave

Plans with `autonomous: false` require user interaction.

Detection: Check autonomous field in frontmatter.

Execution flow for checkpoint plans:

1. Spawn agent for checkpoint plan:

   Task(prompt="{subagent-task-prompt}", subagent_type="gsd-executor", model="{executor_model}")
   

2. Agent runs until checkpoint:

   • Executes auto tasks normally
   • Reaches checkpoint task (e.g., type="checkpoint:human-verify") or auth gate
   • Agent returns with structured checkpoint (see checkpoint-return.md template)

3. Agent return includes (structured format):

   • Completed Tasks table with commit hashes and files
   • Current task name and blocker
   • Checkpoint type and details for user
   • What's awaited from user

4. Orchestrator presents checkpoint to user:

   Extract and display the "Checkpoint Details" and "Awaiting" sections from agent return:

   ## Checkpoint: [Type]
   
   **Plan:** 03-03 Dashboard Layout
   **Progress:** 2/3 tasks complete
   
   [Checkpoint Details section from agent return]
   
   [Awaiting section from agent return]
   

5. User responds:

   • "approved" / "done" → spawn continuation agent
   • Description of issues → spawn continuation agent with feedback
   • Decision selection → spawn continuation agent with choice

6. Spawn continuation agent (NOT resume):

   Use the continuation-prompt.md template:

   Task(
     prompt=filled_continuation_template,
     subagent_type="gsd-executor",
     model="{executor_model}"
   )
   

   Fill template with:

   • {completed_tasks_table}: From agent's checkpoint return
   • {resume_task_number}: Current task from checkpoint
   • {resume_task_name}: Current task name from checkpoint
   • {user_response}: What user provided
   • {resume_instructions}: Based on checkpoint type (see continuation-prompt.md)

7. Continuation agent executes:

   • Verifies previous commits exist
   • Continues from resume point
   • May hit another checkpoint (repeat from step 4)
   • Or completes plan

8. Repeat until plan completes or user stops

Why fresh agent instead of resume: Resume relies on Copilot Code's internal serialization which breaks with parallel tool calls. Fresh agents with explicit state are more reliable and maintain full context.

Checkpoint in parallel context: If a plan in a parallel wave has a checkpoint:

• Spawn as normal
• Agent pauses at checkpoint and returns with structured state
• Other parallel agents may complete while waiting
• Present checkpoint to user
• Spawn continuation agent with user response
• Wait for all agents to finish before next wave

After all waves complete, aggregate results:

## Phase {X}: {Name} Execution Complete

**Waves executed:** {N}
**Plans completed:** {M} of {total}

### Wave Summary

| Wave | Plans            | Status     |
| ---- | ---------------- | ---------- |
| 1    | plan-01, plan-02 | ✓ Complete |
| CP   | plan-03          | ✓ Verified |
| 2    | plan-04          | ✓ Complete |
| 3    | plan-05          | ✓ Complete |

### Plan Details

1. **03-01**: [one-liner from SUMMARY.md]
2. **03-02**: [one-liner from SUMMARY.md]
   ...

### Issues Encountered

[Aggregate from all SUMMARYs, or "None"]

Verify phase achieved its GOAL, not just completed its TASKS.

Spawn verifier:

Task(
  prompt="Verify phase {phase_number} goal achievement.

Phase directory: {phase_dir}
Phase goal: {goal from ROADMAP.md}

Check must_haves against actual codebase. Create VERIFICATION.md.
Verify what actually exists in the code.",
  subagent_type="gsd-verifier",
  model="{verifier_model}"
)

Read verification status:

grep "^status:" "$PHASE_DIR"/*-VERIFICATION.md | cut -d: -f2 | tr -d ' '

Route by status:

Status	Action
passed	Continue to update_roadmap
human_needed	Present items to user, get approval or feedback
gaps_found	Present gap summary, offer /plan-phase.md {phase} --gaps

If passed:

Phase goal verified. Proceed to update_roadmap.

If human_needed:

## ✓ Phase {X}: {Name} — Human Verification Required

All automated checks passed. {N} items need human testing:

### Human Verification Checklist

{Extract from VERIFICATION.md human_verification section}

---

**After testing:**

- "approved" → continue to update_roadmap
- Report issues → will route to gap closure planning

If user approves → continue to update_roadmap. If user reports issues → treat as gaps_found.

If gaps_found:

Present gaps and offer next command:

## ⚠ Phase {X}: {Name} — Gaps Found

**Score:** {N}/{M} must-haves verified
**Report:** {phase_dir}/{phase}-VERIFICATION.md

### What's Missing

{Extract gap summaries from VERIFICATION.md gaps section}

---

## ▶ Next Up

**Plan gap closure** — create additional plans to complete the phase

`/plan-phase.md {X} --gaps`

<sub>`/clear` first → fresh context window</sub>

---

**Also available:**

- `cat {phase_dir}/{phase}-VERIFICATION.md` — see full report
- `/verify-work.md {X}` — manual testing before planning

User runs /plan-phase.md {X} --gaps which:

1. Reads VERIFICATION.md gaps
2. Creates additional plans (04, 05, etc.) with gap_closure: true to close gaps
3. User then runs /execute-phase.md {X} --gaps-only
4. Execute-phase runs only gap closure plans (04-05)
5. Verifier runs again after new plans complete

User stays in control at each decision point.

Update ROADMAP.md to reflect phase completion:

# Mark phase complete
# Update completion date
# Update status

Check planning config:

If COMMIT_PLANNING_DOCS=false (set in load_project_state):

• Skip all git operations for .gsd/ files
• Planning docs exist locally but are gitignored
• Log: "Skipping planning docs commit (commit_docs: false)"
• Proceed to offer_next step

If COMMIT_PLANNING_DOCS=true (default):

• Continue with git operations below

Commit phase completion (roadmap, state, verification):

git add .gsd/ROADMAP.md .gsd/STATE.md .gsd/phases/{phase_dir}/*-VERIFICATION.md
git add .gsd/REQUIREMENTS.md  # if updated
git commit -m "docs(phase-{X}): complete phase execution"

Present next steps based on milestone status:

If more phases remain:

## Next Up

**Phase {X+1}: {Name}** — {Goal}

`/plan-phase.md {X+1}`

<sub>`/clear` first for fresh context</sub>

If milestone complete:

MILESTONE COMPLETE!

All {N} phases executed.

`/complete-milestone.md`

<context_efficiency> Orchestrator: ~10-15% context (frontmatter, spawning, results). Subagents: Fresh 200k each (full workflow + execution). No polling (Task blocks). No context bleed. </context_efficiency>

<failure_handling> Subagent fails mid-plan:

• SUMMARY.md won't exist
• Orchestrator detects missing SUMMARY
• Reports failure, asks user how to proceed

Dependency chain breaks:

• Wave 1 plan fails
• Wave 2 plans depending on it will likely fail
• Orchestrator can still attempt them (user choice)
• Or skip dependent plans entirely

All agents in wave fail:

• Something systemic (git issues, permissions, etc.)
• Stop execution
• Report for manual investigation

Checkpoint fails to resolve:

• User can't approve or provides repeated issues
• Ask: "Skip this plan?" or "Abort phase execution?"
• Record partial progress in STATE.md </failure_handling>

**Resuming interrupted execution:**

If phase execution was interrupted (context limit, user exit, error):

1. Run /execute-phase.md {phase} again
2. discover_plans finds completed SUMMARYs
3. Skips completed plans
4. Resumes from first incomplete plan
5. Continues wave-based execution

STATE.md tracks:

• Last completed plan
• Current wave
• Any pending checkpoints