Phase 3 Plan

Phase Refine Plan Skill

Role: Domain-agnostic workflow skill for transforming solution outline deliverables into optimized, executable tasks. Loaded by pm-workflow:task-plan-agent.

Key Pattern: Reads deliverables with metadata and profiles list from solution_outline.md, creates one task per profile (1:N mapping), resolves skills from architecture based on module + profile, applies aggregation/split analysis, creates tasks with explicit skill lists.

Contract Compliance

MANDATORY: All tasks MUST follow the structure defined in the central contracts:

CRITICAL - Steps Field:

• The steps field MUST contain file paths from the deliverable's Affected files section
• Steps must NOT be descriptive text (e.g., "Update AuthController.java" is INVALID)
• Validation rejects tasks with non-file-path steps

Input

Output

status: success | error
plan_id: {echo}
optimization_summary:
  deliverables_processed: N
  tasks_created: M
  aggregations: N
  splits: N
  parallelizable_groups: N
tasks_created[M]: {number, title, deliverables, depends_on}
execution_order: {parallel groups}
message: {error message if status=error}

Workflow

Step 1: Load All Deliverables

Read the solution document to get all deliverables with metadata:

python3 .plan/execute-script.py pm-workflow:manage-solution-outline:manage-solution-outline \
  list-deliverables \
  --plan-id {plan_id}

For each deliverable, extract:

• metadata.change_type, metadata.execution_mode
• metadata.domain (single value)
• metadata.module (module name from architecture)
• metadata.depends
• profiles (list: implementation, testing)
• affected_files
• verification

Step 2: Build Dependency Graph

Parse depends field for each deliverable:

• Identify independent deliverables (depends: none)
• Identify dependency chains
• Detect cycles (INVALID - reject with error)

Step 3: Analyze for Aggregation

For each pair of deliverables, check if they can be aggregated:

• Same change_type?
• Same domain?
• Same profiles list?
• Same execution_mode (must be automated)?
• Combined file count < 10?
• NO dependency between them? (CRITICAL - cannot aggregate if one depends on the other)

Step 4: Analyze for Splits

For each deliverable, check for split requirements:

• execution_mode: mixed → MUST split
• File count > 15 → CONSIDER splitting

Note: Multiple profiles in **Profiles:** block naturally create multiple tasks (1:N) - no additional splitting needed for profile differences.

Step 5: Create Tasks from Profiles (1:N Mapping)

For each deliverable, create one task per profile in its profiles list:

For each deliverable D:
  1. Query architecture: module --name {D.module}
  For each profile P in D.profiles:
    2. Extract skills: module.skills_by_profile.{P}
    3. Create task with profile P and resolved skills
    4. If P = testing, add depends on implementation task

Query architecture:

python3 .plan/execute-script.py plan-marshall:analyze-project-architecture:architecture \
  module --name {deliverable.module} \
  --trace-plan-id {plan_id}

1:N Task Creation Flow:

solution_outline.md                        TASK-*.toon (created by task-plan)
┌────────────────────────────┐             ┌────────────────────────────┐
│ **Metadata:**              │             │ TASK-001-IMPL              │
│ - domain: java             │             │ profile: implementation    │
│ - module: auth-service     │  ───────►   │ skills: [java-core,        │
│                            │  (1:N)      │          java-cdi]         │
│ **Profiles:**              │             ├────────────────────────────┤
│ - implementation           │  ───────►   │ TASK-002-TEST              │
│ - module_testing           │             │ profile: module_testing    │
│                            │             │ skills: [java-core,        │
└────────────────────────────┘             │          junit-core]       │
                                           │ depends: TASK-001-IMPL     │
                                           └────────────────────────────┘

Log skill resolution (for each task created):

python3 .plan/execute-script.py plan-marshall:manage-logging:manage-log \
  work {plan_id} INFO "[SKILL] (pm-workflow:phase-3-plan) Resolved skills for TASK-{N} from {module}.{profile}: [{task.skills}]"

Aggregation Rule: When aggregating multiple deliverables, they must have same profiles list. Merge resolved skills arrays (union) per profile.

Step 6: Create Optimized Tasks

For aggregated deliverables or single deliverables, create tasks using heredoc:

python3 .plan/execute-script.py pm-workflow:manage-tasks:manage-tasks add \
  --plan-id {plan_id} <<'EOF'
title: {aggregated title}
deliverables: [{n1}, {n2}, {n3}]
domain: {domain from deliverable}
profile: {profile from deliverable}
phase: execute
description: |
  {combined description}

steps:
  - {file1}
  - {file2}
  - {file3}

depends_on: TASK-1, TASK-2

skills:
  - {skill1 from architecture}
  - {skill2 from architecture}

verification:
  commands:
    - {cmd1}
  criteria: {criteria}
EOF

MANDATORY - Log each task creation:

python3 .plan/execute-script.py plan-marshall:manage-logging:manage-log \
  work {plan_id} INFO "[ARTIFACT] (pm-workflow:phase-3-plan) Created TASK-{N}: {title}"

Key Fields:

• domain: Single domain from deliverable
• profile: implementation or testing (determines workflow skill at execution)
• skills: Domain skills only (system skills loaded by agent)
• steps: File paths from Affected files (NOT descriptive text)

Step 7: Determine Execution Order

Compute parallel execution groups:

execution_order:
  parallel_group_1: [TASK-1, TASK-3]    # No dependencies
  parallel_group_2: [TASK-2, TASK-4]    # Both depend on group 1
  parallel_group_3: [TASK-5]            # Depends on group 2

Parallelism rules:

• Tasks with no depends_on go in first group
• Tasks depending on same prior tasks can run in parallel
• Sequential dependencies remain sequential

Step 8: Record Issues as Lessons

On ambiguous deliverable or planning issues:

python3 .plan/execute-script.py plan-marshall:manage-lessons:manage-lesson add \
  --component-type skill \
  --component-name task-plan \
  --category observation \
  --title "{issue summary}" \
  --detail "{context and resolution approach}" \
  --trace-plan-id {plan_id}

Step 9: Return Results

Output:

status: success
plan_id: {plan_id}

optimization_summary:
  deliverables_processed: {N}
  tasks_created: {M}
  aggregations: {count of deliverable groups}
  splits: {count of split deliverables}
  parallelizable_groups: {count of independent task groups}

tasks_created[M]{number,title,deliverables,depends_on}:
1,Implement UserService,[1],none
2,Implement UserRepository,[2],none
3,Add integration tests,[3],"TASK-1" "TASK-2"

execution_order:
  parallel_group_1: [TASK-1, TASK-2]
  parallel_group_2: [TASK-3]

lessons_recorded: {count}

Optimization Decision Table

Note: Multiple profiles in one deliverable naturally create multiple tasks (1:N). This is not a "split" - it's the standard task creation pattern.

Dependency Rules for Aggregation

Skill Resolution Guidelines

Skills are resolved from architecture based on module + profile:

Error Handling

Circular Dependencies

If deliverable dependencies form a cycle:

• Error: "Circular dependency detected: D1 -> D2 -> D1"
• Do NOT create tasks

Module Not in Architecture

If deliverable.module is not found in architecture:

• Error: "Module '{module}' not found in architecture - run architecture discovery"
• Record as lesson learned

Profile Not in Module

If a profile from deliverable.profiles is not in module.skills_by_profile:

• Error: "Profile '{profile}' not found in {module}.skills_by_profile"
• Record as lesson learned

Ambiguous Deliverable

If deliverable metadata incomplete:

• Generate task with defaults
• Add lesson-learned for future reference
• Note ambiguity in task description

Integration

Invoked by: pm-workflow:task-plan-agent (thin agent)

Script Notations (use EXACTLY as shown):

• pm-workflow:manage-solution-outline:manage-solution-outline - Read deliverables (list-deliverables, read)
• plan-marshall:analyze-project-architecture:architecture - Resolve skills (module --name {module})
• pm-workflow:manage-tasks:manage-tasks - Create tasks (add --plan-id X <<'EOF' ... EOF)
• plan-marshall:manage-lessons:manage-lesson - Record lessons on issues (add)

Consumed By:

• pm-workflow:task-execute-agent - Reads tasks and executes them