Claude Code Plugins

Community-maintained marketplace

Feedback

Adaptive brainstorming for plugin concepts and improvements when exploring ideas, not implementing. Autonomous triggers: "I want to make...", "Explore improvements to...", "brainstorm", "ideate", "new plugin idea", "what if I added". NOT for implementing existing improvement proposals (use plugin-improve skill).

Install Skill

Shared

Installs to .agents/skills, used by Codex, Amp, Warp, Cursor, OpenCode, and more.

CodexAmp
Warp
CursorOpenCode
Cline
Gemini CLI
GitHub Copilot
Personal

Available across projects.

$npx skills-installer add @majiayu000/claude-skill-registry/plugin-ideation --client shared
Project

Writes to .agents/skills.

$npx skills-installer add @majiayu000/claude-skill-registry/plugin-ideation -p --client shared
Note: Review the skill instructions before using it.

SKILL.md

name plugin-ideation
description Adaptive brainstorming for plugin concepts and improvements when exploring ideas, not implementing. Autonomous triggers: "I want to make...", "Explore improvements to...", "brainstorm", "ideate", "new

name: plugin-ideation description: Adaptive brainstorming for plugin concepts and improvements when exploring ideas, not implementing. Autonomous triggers: "I want to make...", "Explore improvements to...", "brainstorm", "ideate", "new plugin idea", "what if I added". NOT for implementing existing improvement proposals (use plugin-improve skill). allowed-tools: - Read - Write - Bash preconditions: - None (entry point)

plugin-ideation Skill

Purpose: Interactive brainstorming and creative vision capture for new plugins and improvements to existing plugins.

This skill uses mode-based workflow loading to optimize context window usage. After detecting which mode applies (New Plugin vs Improvement), the appropriate workflow reference is loaded on-demand.

Mode Detection

mode_detection must_complete_before_proceeding **Check if plugin exists:**
grep "^### $PLUGIN_NAME$" PLUGINS.md

Route based on result:

  • If found → Load Improvement Mode workflow
  • If not found → Load New Plugin Mode workflow

Workflow References

After mode detection completes, load the appropriate workflow:

New Plugin Mode

When to use: Plugin does not exist in PLUGINS.md (new plugin ideation)

Workflow: See new-plugin-workflow.md

Overview: 8-phase workflow from free-form collection through creative brief creation

  • Phase 1: Free-form collection
  • Phase 2: Gap analysis (4-tier priority system)
  • Phase 3: Question batch generation (adaptive 4-question batches)
  • Phase 4: Decision gate (finalize or iterate)
  • Phase 5: Plugin name validation
  • Phase 6: Document creation (creative-brief.md + PLUGINS.md)
  • Phase 7: Session handoff (.continue-here.md)
  • Phase 8: Decision menu with parallel workflow option

Key features:

  • Adaptive questioning based on tier-based gap analysis
  • Context accumulation across question batches
  • Parallel workflow path (Phase 8.1) for faster iteration
  • Continuous iteration support until user says "finalize"

Improvement Mode

When to use: Plugin exists in PLUGINS.md (improvement brainstorming)

Workflow: See improvement-workflow.md

Overview: 7-phase workflow from vagueness detection through improvement proposal creation

  • Phase 0: Vagueness detection (brainstorm vs implement routing)
  • Phase 1: Free-form collection
  • Phase 2: Gap analysis (3-tier priority system)
  • Phase 3: Question batch generation
  • Phase 4: Decision gate (finalize or iterate)
  • Phase 5: Document creation (improvement proposal)
  • Phase 6: Session handoff (.continue-here.md)
  • Phase 7: Decision menu

Key features:

  • Vagueness detection to route between ideation and implementation
  • 3-tier gap analysis specific to improvements
  • Version impact tracking (PATCH/MINOR/MAJOR)
  • Backward compatibility consideration

Mode-Based Loading Instructions

  1. Complete mode detection first (check PLUGINS.md)
  2. Load appropriate workflow reference:
  3. Execute loaded workflow following all phase instructions
  4. Follow checkpoint protocol at decision gates

This structure reduces initial context load by ~50% (from 8k to 4k tokens) while preserving full functionality through on-demand reference loading.