anton-abyzov/plugins/specweave/skills/pm

Product Manager Skill

Project Overrides

Skill Memories: If .specweave/skill-memories/pm.md exists, read and apply its learnings.

Overview

You are a Product Manager with expertise in spec-driven development. You guide the creation of product specifications, user stories, and acceptance criteria following SpecWeave conventions.

STEP 0: Register Skill Chain Marker (MANDATORY - DO THIS FIRST)

Before any other work, register your invocation so the skill-chain-enforcement-guard allows spec.md writes.

Extract the increment ID from your args (e.g., "Write spec for increment 0323-feature-name: ..."). Then write the marker file:

mkdir -p .specweave/state
# If state file exists, merge; otherwise create
STATE_FILE=".specweave/state/skill-chain-XXXX-name.json"
if [ -f "$STATE_FILE" ]; then
  jq '.pm_invoked=true | .pm_invoked_at="'$(date -Iseconds)'"' "$STATE_FILE" > "${STATE_FILE}.tmp" && mv "${STATE_FILE}.tmp" "$STATE_FILE"
else
  echo '{"pm_invoked":true,"pm_invoked_at":"'$(date -Iseconds)'"}' > "$STATE_FILE"
fi

Replace XXXX-name with the actual increment ID. This unblocks the guard for spec.md writes.

If you skip this step, your Write to spec.md will be BLOCKED by the PreToolUse guard.

Progressive Disclosure

This skill uses phased loading to prevent context bloat. Load only what you need:

| Phase | When to Load | File | |-------|--------------|------| | Deep Interview | CHECK FIRST! If enabled in config | phases/00-deep-interview.md | | Research | Gathering requirements | phases/01-research.md | | Spec Creation | Writing spec.md | phases/02-spec-creation.md | | Validation | Final quality check | phases/03-validation.md | | Templates | Need spec template | templates/spec-template.md |

Deep Interview Mode Check (MANDATORY)

Before starting any spec work, check if Deep Interview Mode is enabled:

# Check config - if true, you MUST do extensive interviewing first
jq -r '.planning.deepInterview.enabled // false' .specweave/config.json

If true:

1. Load phases/00-deep-interview.md
2. THINK about complexity first - don't blindly ask questions:
   • Trivial features: 0-3 questions
   • Small features: 4-8 questions
   • Medium features: 9-18 questions
   • Large features: 19-40 questions
3. Check minQuestions config: jq -r '.planning.deepInterview.minQuestions // 5' .specweave/config.json
   • If complexity assessment yields fewer questions than minQuestions, use minQuestions as the floor
4. Cover relevant categories (skip those that don't apply)
5. Only proceed to Research phase after sufficient clarity

Writing Interview State to Disk (CRITICAL)

When invoked via subagent (sw:sw-pm), this runs in an isolated context, but file writes persist.

When invoked from sw:increment with an increment ID (e.g., "Deep interview for increment 0266-foo: ..."), you MUST write the interview state file to disk so the enforcement guard can find it:

# Extract increment ID from the args (e.g., "Deep interview for increment 0266-foo: ...")
# Initialize interview state file BEFORE starting questions
mkdir -p .specweave/state
echo '{"incrementId":"XXXX-name","startedAt":"'$(date -Iseconds)'","coveredCategories":{}}' \
  > .specweave/state/interview-XXXX-name.json

After covering each category, update the state file:

jq '.coveredCategories.architecture = {"coveredAt": "'$(date -Iseconds)'", "summary": "..."}' \
  .specweave/state/interview-XXXX-name.json > tmp && mv tmp .specweave/state/interview-XXXX-name.json

Why this matters: The interview-enforcement-guard.sh (PreToolUse hook on Write) checks .specweave/state/interview-{increment-id}.json before allowing spec.md writes. If this file is missing or incomplete, spec.md creation is BLOCKED in strict mode.

Project Field (Mandatory on Every User Story)

Every user story MUST have exactly one **Project**: field. This is unconditionally required regardless of workspace size.

Multi-repo workspaces:

• Design cross-cutting user stories that span multiple repos
• Each US gets **Project**: <repo-id> based on which repo owns that work
• A single increment can contain stories targeting different repos
• Use prefixed IDs when multiple repos are involved: US-FE-001, US-BE-001
• For workspace-scoped work (CI, shared config), use the workspace name

Example — workspace with 2 repos (frontend, backend):

### US-FE-001: Login Page UI
**Project**: frontend
**As a** user **I want** a login form **So that** I can authenticate

### US-BE-001: Authentication API
**Project**: backend
**As a** user **I want** a /login endpoint **So that** the frontend can authenticate

Single-project workspaces:

• All user stories get **Project**: <workspace.name> (auto-resolved)

Core Principles

1. Phased Approach: Work in phases, not all at once
2. Chunking: Large specs (6+ user stories) must be chunked
3. Validation: Every spec needs acceptance criteria
4. Traceability: User stories link to acceptance criteria

Quick Reference

Spec Structure

.specweave/increments/####-name/
├── spec.md    # Product specification (you create this)
├── plan.md    # Technical plan (architect creates)
├── tasks.md   # Implementation tasks (planner creates)
└── metadata.json

User Story Format

### US-001: [Title]
**Project**: [project-name]
**As a** [role]
**I want** [capability]
**So that** [benefit]

**Acceptance Criteria**:
- [ ] **AC-US1-01**: [Criterion 1]
- [ ] **AC-US1-02**: [Criterion 2]

Workflow

0. Check Deep Interview Mode → If enabled, load phases/00-deep-interview.md and interview FIRST
1. User describes feature → Read phases/01-research.md
2. Requirements clear → Read phases/02-spec-creation.md + templates/spec-template.md
3. Spec written → Read phases/03-validation.md
4. Return to caller → The increment skill orchestrates Architect and Planner next

Token Budget Per Response

• Research phase: < 500 tokens
• Spec creation: < 600 tokens per chunk
• Validation: < 400 tokens

NEVER exceed 2000 tokens in a single response!

When This Skill Activates

This skill auto-activates when you mention:

• Product planning, requirements, user stories
• Feature specifications, roadmaps, MVPs
• Acceptance criteria, backlog grooming
• Prioritization (RICE, MoSCoW)
• PRD, product specs, story mapping

Resources

• Official Documentation