Claude-skill-registry backlog-scan

Bulk backlog scanner that analyzes the entire finans codebase vs CLAUDE.md, identifies ALL gaps, and generates a comprehensive, prioritized, numbered task backlog. Uses broad→narrow pattern.

install
source · Clone the upstream repo
git clone https://github.com/majiayu000/claude-skill-registry
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/backlog-scan" ~/.claude/skills/majiayu000-claude-skill-registry-backlog-scan && rm -rf "$T"
manifest: skills/data/backlog-scan/SKILL.md
source content

Backlog Scan

Autonomous bulk scanner for comprehensive backlog generation. Scans entire codebase against CLAUDE.md spec, finds ALL gaps, generates numbered tasks.

Pattern: Broad→Narrow (scan everything, then focus on gaps)

🚨 PLANNING ONLY - NO IMPLEMENTATION This skill creates task files. It does NOT write code or start implementation.

Use

/new-task
instead for a single specific task (narrow→broad pattern).

When to Use This Skill

Use this skill when:

  • Starting a new development phase - need comprehensive backlog
  • Backlog is stale or empty
  • Want to discover ALL missing features, refactors, improvements
  • Need to organize and prioritize existing unnumbered tasks
  • Running
    /discover-tasks
    slash command

DO NOT use this skill for:

  • Creating a single specific task → use
    feature-planning
    skill
  • Implementing tasks → this skill only discovers and plans
  • Quick bug fixes → just fix them directly

Context Loading (ALWAYS FIRST)

Before any analysis, load ALL context:

┌─────────────────────────────────┐
│        CONTEXT LOADING          │
│  • ALL rules from .claude/rules/│
│  • ALL done/ tasks (recent 30)  │
│  • ALL available skills         │
│  • CLAUDE.md completely         │
└─────────────────────────────────┘
  1. Load ALL rules from

    .claude/rules/

    • List:
      find .claude/rules -name "*.md"
    • Read rules for ALL areas (frontend/, backend/, components/, e2e/)
    • Rules contain patterns, gotchas, decisions
  2. Check ALL completed tasks in

    .task-board/done/

    • Scan recent 20-30 completed tasks
    • Learn task format and quality standards
    • Identify recurring themes
  3. Note ALL available skills

    • frontend-design
      - UI/CSS implementations
    • node-backend
      - API routes
    • rule-making-skill
      - Create missing rules
    • storybook-stories
      - Component documentation
    • task-board
      - Detailed task planning

Scan Workflow

Phase 1: Full Codebase Analysis

Goal: Understand EVERYTHING that exists

  1. Read CLAUDE.md completely

    • ALL described features, architecture, requirements
    • Technology stack and patterns
    • Finans domain (portfolio tracking, F.I.R.E., calculators)
  2. Scan ALL codebase areas

    • /frontend/src/features/
      - existing features
    • /backend/src/routes/
      - API routes
    • /components/src/
      - component library
    • /e2e/
      - test coverage
    • TODOs, FIXMEs in code
  3. Review ALL existing tasks

    • .task-board/backlog/
    • .task-board/in-progress/
    • .task-board/on-hold/
    • Identify gaps, duplicates, vague tasks

Phase 2: Gap Analysis

Goal: Find ALL gaps between spec and implementation

  • Features in CLAUDE.md NOT yet implemented
  • Missing infrastructure (CI/CD, monitoring)
  • Code quality issues (validation, error handling)
  • Technical debt (deprecated patterns, TODOs)
  • Finans-specific gaps (Norwegian formatting, calculators)

Phase 3: Task Generation

Quality Bar (ALL must be met):

  • ✅ Clear value: Obvious user benefit OR technical necessity
  • ✅ Well-scoped: Not epic-sized, not trivial
  • ✅ Actionable: Can implement without major unknowns
  • ✅ Domain-aligned: Fits finans portfolio tracking
  • ✅ Non-redundant: Not covered by existing task

For each gap that meets quality bar:

  1. Invoke
    task-board
    skill for detailed planning
  2. Provide context and priority
  3. Review output, assign number

Phase 4: Numbering & Organization

🚨 CRITICAL: Scan ALL folders for highest number:

1. Glob: .task-board/**/*.md
2. Scan: backlog/ + in-progress/ + done/
3. Find highest number
4. Next task = highest + 1

Never reuse numbers - done/ tasks are immutable history.

File format:

NNN-TYPE-description.md

  • 076-FEATURE-portfolio-dashboard.md
  • 077-REFACTOR-validation-consolidation.md

Phase 5: Cleanup

  • Enhance vague existing tasks
  • Merge duplicates
  • Remove outdated tasks
  • Final state: numbered, no duplicates, quality bar met

Output

Summary report with:

  • Tasks created/enhanced/merged
  • Breakdown by type and priority
  • Top 5 priorities
  • Next steps

Delegates To

  • task-board
    skill - for detailed plan files

See Also

  • feature-planning
    skill - for single task (narrow→broad)
  • .task-board/WORKFLOW.md
    - task management workflow
  • .claude/CLAUDE.md
    - project instructions