/mega-approve

description: "Approve the mega-plan and start feature execution. Creates worktrees and generates PRDs for each feature. Usage: /plan-cascade:mega-approve [--flow <quick|standard|full>] [--tdd <off|on|auto>] [--confirm] [--no-confirm] [--spec <off|auto|on>] [--first-principles] [--max-questions N] [--auto-prd] [--agent <name>] [--prd-agent <name>] [--impl-agent <name>]"

Approve Mega Plan and Start Execution

Approve the mega-plan and begin executing features in **batch-by-batch** order with **FULL AUTOMATION**.

Execution Flow Parameters

This command accepts flow control parameters that propagate to ALL feature executions:

Parameter Priority

Parameters are resolved in the following order when multiple sources exist:

1. **Command-line flags to THIS command** (highest priority)

• Example: `/plan-cascade:mega-approve --flow full --tdd on`
• Overrides all other sources

2. **mega-plan.json configuration** (saved by `/plan-cascade:mega-plan`)

• `flow_config`, `tdd_config`, `spec_config`, `execution_config`
• Used when no command-line flag is provided

3. **Default values** (lowest priority)

**Priority Resolution Example:**

# Scenario 1: Command-line overrides mega-plan
mega-plan.json contains: flow_config.level = "standard"
You run: /plan-cascade:mega-approve --flow full
Result: All features use FLOW = "full"

# Scenario 2: Use mega-plan config
mega-plan.json contains: tdd_config.mode = "on"
You run: /plan-cascade:mega-approve
Result: All features use TDD = "on"

# Scenario 3: Partial override
mega-plan.json contains: flow="full", tdd="on"
You run: /plan-cascade:mega-approve --flow standard
Result: All features use FLOW = "standard", TDD = "on"

**Parameter Propagation to Features:**

• Resolved parameters are passed to PRD generation agents via prompt
• PRD generation agents write these parameters to each feature's `prd.json`
• Story execution agents read parameters from their feature's `prd.json`

**Special Case - Spec Interview Parameters:**

• `--spec`, `--first-principles`, `--max-questions` are orchestrator-only
• Used in Step 6.0 (this command) to run spec interviews
• NOT propagated to feature agents (see Step 2.0.2 notes)

`--flow <quick|standard|full>`

Override the execution flow depth for all feature approve phases.

| Flow | Gate Mode | AI Verification | Code Review | Test Enforcement | |------|-----------|-----------------|-------------|------------------| | `quick` | soft | disabled | no | no | | `standard` | soft | enabled | no | no | | `full` | **hard** | enabled | **required** | **required** |

**FULL Flow Enforcement** (propagates to all features):

• Quality gates BLOCK execution on failure
• Code review is REQUIRED after each story
• Test file changes are REQUIRED alongside code changes
• With `--confirm`: Batch-level confirmation before each feature batch (see below)

`--tdd <off|on|auto>`

Control Test-Driven Development mode for all feature story executions.

| Mode | Description | |------|-------------| | `off` | TDD disabled | | `on` | TDD enabled with prompts and compliance checks | | `auto` | Automatically enable TDD for high-risk stories (default) |

`--confirm`

Require explicit user confirmation before starting each **feature batch**.

**IMPORTANT: Batch-Level Confirmation Design**

In mega-plan execution, confirmation happens at the **batch level**, not inside sub-agents:

Batch 1: [feature-A, feature-B, feature-C]
    ↓
[CONFIRMATION POINT] ← User confirms here (Step 4.5)
    ↓
All 3 features execute in parallel (no further confirmation)
    ↓
Batch 2: [feature-D, feature-E]
    ↓
[CONFIRMATION POINT] ← User confirms here
    ↓
...

Why batch-level confirmation:

• Sub-agents run in background (`run_in_background=true`)
• Background agents cannot use `AskUserQuestion`
• Multiple parallel agents requesting confirmation would be chaotic
• Batch-level provides oversight while preserving parallelism

`--no-confirm`

Disable batch confirmation, even when using FULL flow.

• Overrides `--confirm` flag
• Overrides mega-plan.json's `execution_config.require_batch_confirm`
• Overrides FULL flow's default confirmation behavior
• Useful for CI pipelines that want strict quality gates without interactive confirmation

`--spec <off|auto|on>`

Run a planning-time **spec interview** per feature (in the orchestrator only) to produce `spec.json/spec.md`, then compile to `prd.json` before story execution starts.

• `auto` (default): enabled when `--flow full`, otherwise disabled
• `on`: always run interview before PRD finalization
• `off`: never run interview

`--first-principles`

Enable first-principles questions (only when spec interview runs).

`--max-questions N`

Soft cap for interview length (recorded in `.state/spec-interview.json` per feature).

Path Storage Modes

This command works with both new and legacy path storage modes:

New Mode (Default)

Files are stored in user data directory:

• **Windows**: `%APPDATA%/plan-cascade/<project-id>/`
• **Unix/macOS**: `~/.plan-cascade/<project-id>/`

File locations:

• `mega-plan.json`: `<user-dir>/mega-plan.json`
• `.mega-status.json`: `<user-dir>/.state/.mega-status.json`
• Worktrees: `<user-dir>/.worktree/<feature-name>/`

Legacy Mode

All files in project root:

• `mega-plan.json`: `<project-root>/mega-plan.json`
• `.mega-status.json`: `<project-root>/.mega-status.json`
• Worktrees: `<project-root>/.worktree/<feature-name>/`

The command uses PathResolver to determine correct paths automatically.

Multi-Agent Collaboration

Mega-plan execution supports multiple AI agents at two levels:

1. **PRD Generation**: Agent that generates PRDs for each feature worktree 2. **Story Execution**: Agent that executes stories within each feature

Supported Agents

| Agent | Type | Best For | |-------|------|----------| | `claude-code` | task-tool | General purpose (default, always available) | | `codex` | cli | PRD generation, bug fixes | | `aider`