本文へ移動
cccskills
無料GitHub で公開

feature-manifest

Manage feature manifests for code traceability. Use when creating new features, updating existing features, checking feature health, or exploring the feature-to-code relationship. Activates for manifest validation, feature creation, changelog updates, and traceability queries.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md6.5 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Feature Manifest Management

This skill manages the feature manifest system that tracks the relationship between features and their implementations.

DECISION POINTS

When working with code changes, use this decision tree:

Is this code change happening?
├── Creating new feature directory/component
│   ├── New feature has >5 files OR touches multiple domains
│   │   └── → CREATE new manifest (npm run feature:create)
│   └── Small utility/helper (<5 files, single purpose)
│       └── → ADD to existing related manifest
├── Modifying existing tracked files
│   ├── Files belong to known feature (check with: npm run feature:info -- --files <path>)
│   │   ├── Adding/removing files to feature
│   │   │   └── → UPDATE manifest files list + last_modified date
│   │   └── Just changing file contents
│   │       └── → UPDATE last_modified date only
│   └── Files are orphaned (not in any manifest)
│       └── → Determine owner with npm run feature:health, then UPDATE appropriate manifest
└── Removing/deprecating code
    ├── Entire feature being removed
    │   └── → SET status to 'deprecated', add deprecation changelog entry
    └── Some files being removed from feature
        └── → UPDATE manifest files list, add changelog entry

For manifest validation failures:

  • Missing files → Add files to manifest or remove from filesystem
  • Extra files → Remove from manifest files list
  • Stale manifest (>90 days) → Update last_modified and add changelog entry
  • No tests listed → Add test files or mark as technical debt

FAILURE MODES

1. Orphan File Syndrome

  • Detection: npm run feature:health shows files in "Orphaned Files" section
  • Cause: Files created without updating any manifest
  • Fix: Run npm run feature:info -- --files <path> to find logical owner, then add files to appropriate manifest

2. Ghost Manifest References

  • Detection: Validation errors about missing files that were deleted
  • Cause: Files removed from codebase but not from manifest
  • Fix: Edit manifest to remove deleted files from implementation.files, update last_modified, add changelog entry

3. Stale Dependency Web

  • Detection: Features referencing deprecated or renamed dependencies in dependencies.internal
  • Cause: Feature refactoring without updating dependent manifests
  • Fix: Search codebase for references to old feature ID, update all dependent manifests

4. Test Coverage Blind Spot

  • Detection: npm run feature:health shows features in "Without Tests" section
  • Cause: Tests exist but not listed in manifest, or no tests written
  • Fix: If tests exist, add to appropriate test section; if missing, create tests or document as technical debt

5. Manifest Explosion

  • Detection: Too many tiny manifests (>20 manifests for <100 files)
  • Cause: Creating manifests for every small utility
  • Fix: Consolidate related small features into logical groupings, deprecate overly granular manifests

WORKED EXAMPLES

Example: Adding authentication middleware to existing API feature

  1. Identify impact: New middleware affects multiple API routes

    npm run feature:info -- --files src/app/api
    # Shows: api-core.yaml owns most API files
    
  2. Check current manifest:

    npm run feature:info -- api-core
    # Review current files list and dependencies
    
  3. Decision: This is modifying existing feature (middleware is part of API infrastructure)

  4. Update manifest: Add new middleware files to api-core.yaml:

    implementation:
      files:
        - src/app/api/middleware/auth.ts  # NEW
        - src/lib/auth-utils.ts          # NEW
        # ... existing files
    
  5. Update metadata:

    dependencies:
      internal:
        - features/user-management.yaml  # NEW dependency
    history:
      last_modified: "2024-12-23"       # TODAY
    changelog:
      - version: "1.2.0"                # NEW entry
        date: "2024-12-23"
        changes:
          - "Added authentication middleware for API routes"
    
  6. Validate: npm run feature:validate confirms no errors

What novice misses: Updating dependencies.internal when middleware uses user auth What expert catches: All affected API routes need the auth dependency listed

QUALITY GATES

Task is complete when ALL conditions are met:

  • npm run feature:validate passes with no errors
  • npm run feature:health shows no new orphaned files
  • All implementation files listed in manifest exist in codebase
  • All files in manifest have been modified more recently than last_modified date OR manifest updated
  • If adding dependencies, all internal deps reference existing manifests
  • If modifying feature, changelog has new entry with today's date
  • If creating feature, manifest has at least entry_point and one implementation file
  • Status field matches reality (complete features have tests, deprecated features noted)
  • Any new environment variables are documented in dependencies.env_vars
  • Database tables/API routes are documented if feature touches them

NOT-FOR BOUNDARIES

Do NOT use this skill for:

  • Archived/legacy code → Use git-archaeology skill instead for historical analysis
  • Third-party vendor files (node_modules, .next, etc.) → These are build artifacts, not features
  • Configuration files (package.json, tsconfig.json, etc.) → Use project-config skill instead
  • Documentation-only changes (README, docs/) → Use documentation skill instead
  • Build/deployment scripts → Use devops-automation skill instead
  • Test files that don't belong to features (setup files, mocks) → Use test-infrastructure skill instead

For cross-cutting concerns spanning multiple features:

  • Use architecture-analysis skill to understand system boundaries first
  • Then update multiple manifests consistently using this skill

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Expert in 2000s-era music visualization (Milkdrop, AVS, Geiss) and modern WebGL implementations. Specializes in Butterchurn integration, Web Audio API AnalyserNode FFT data, GLSL shaders for audio-reactive visuals, and psychedelic generative art. Activate on "Milkdrop", "music visualization", "WebGL visualizer", "Butterchurn", "audio reactive", "FFT visualization", "spectrum analyzer". NOT for simple bar charts/waveforms (use basic canvas), video editing, or non-audio visuals.

日本語の概要は準備中です。原文の説明を表示しています。

curiositech/port-daddy22026年10月8日 更新

Expert legal research agent for finding and scraping expungement data state by state. Knows authoritative sources, URL patterns, Firecrawl configuration, and 2026 legal landscape.

日本語の概要は準備中です。原文の説明を表示しています。

curiositech/port-daddy22026年10月8日 更新

Expert in 3D computer vision labeling tools, workflows, and AI-assisted annotation for LiDAR, point clouds, and sensor fusion. Covers SAM4D/Point-SAM, human-in-the-loop architectures, and vertical-specific training strategies. Activate on '3D labeling', 'point cloud annotation', 'LiDAR labeling', 'SAM 3D', 'SAM4D', 'sensor fusion annotation', '3D bounding box', 'semantic segmentation point cloud'. NOT for 2D image labeling (use clip-aware-embeddings), general ML training (use ml-engineer), video annotation without 3D (use computer-vision-pipeline), or VLM prompt engineering (use prompt-engineer).

日本語の概要は準備中です。原文の説明を表示しています。

curiositech/port-daddy22026年10月8日 更新

Apply crisis decision-making research to agent routing, uncertainty triage, and coordination failure analysis in time-pressured systems. Use when diagnosing handoff failures, analytical paralysis, or expert judgment under incomplete information. NOT for routine coding, simple CRUD design, or static single-agent tasks with complete information.

日本語の概要は準備中です。原文の説明を表示しています。

curiositech/port-daddy22026年10月8日 更新

Use for insight, reframing, contradiction, impasse, and anomaly-driven problem solving when execution effort no longer helps. NOT for routine optimization, error correction, or well-specified tasks with known solution paths.

日本語の概要は準備中です。原文の説明を表示しています。

curiositech/port-daddy22026年10月8日 更新

Apply cognitive task analysis to expert work that depends on perceptual cues, branching judgment, and recurring monitoring loops. Use when decomposing expert capability into agent structure, simulation design, or validation interviews. NOT for ordinary step-by-step SOP capture or simple pipelines with no tacit cue layer.

日本語の概要は準備中です。原文の説明を表示しています。

curiositech/port-daddy22026年10月8日 更新

curiositech のスキルをすべて見る

このスキルの問題を報告する