---

name: idea-storer description: Integrate with IDEA Storer CLI to read, manage, and act on stored development ideas. Use when user mentions "ideas", "IS", "idea storer", "check my ideas", "what should I work on", or asks about brainstorming and task planning.

IDEA Storer (IS) Integration

A lightning-fast CLI tool for capturing development ideas that AI coding agents can understand and use.

Quick Reference

Discovery

Check for ideas in this order:

1. Project-specific: .ideas.json in current directory or git root
2. Global fallback: ~/.config/ideastorer/ideas.json (Linux/macOS) or %APPDATA%\ideastorer\ideas.json (Windows)

Reading Ideas

When the user mentions ideas, tasks, or asks "what should I work on":

1. Read the .ideas.json file using available file reading tools
2. Parse the JSON and extract active ideas
3. Present relevant ideas based on current context
4. Suggest which ideas relate to current work

JSON Schema

Copy{
  "version": "1.0.0",
  "ideas": [
    {
      "id": "abc12345",
      "content": "The idea text",
      "tags": ["tag1", "tag2"],
      "priority": "low|medium|high",
      "status": "active|done|archived",
      "created_at": "ISO8601",
      "updated_at": "ISO8601",
      "context": {
        "project": "project-name",
        "directory": "/path/to/project",
        "git_branch": "branch-name"
      }
    }
  ],
  "metadata": {
    "last_modified": "ISO8601",
    "idea_count": 1
  }
}

CLI Commands

Copy# Quick capture
is "my brilliant idea"           # Add idea instantly
is "fix bug" -t bug,urgent       # Add with tags
is "critical task" -p high       # Add with priority

# Viewing
is list                          # Show recent ideas
is list -n 10                    # Show 10 ideas
is list -t feature               # Filter by tag
is list --status done            # Filter by status
is list --all                    # Show all ideas

# Searching
is search "authentication"       # Search by content
is search "api" -t backend       # Search with tag filter

# Managing
is done abc123                   # Mark as completed (partial ID works)
is delete abc123                 # Remove idea
is delete abc123 --force         # Remove without confirmation

# Exporting
is export                        # Export as JSON
is export --format markdown      # Export as Markdown
is export -o backup.json         # Export to file

# Setup
is init                          # Initialize .ideas.json
is --global init                 # Initialize global storage

Workflow Integration

Before Starting Work

When user asks "What should I work on?" or "Check my ideas":

1. Read .ideas.json from project root
2. Filter for status: "active" ideas
3. Sort by priority (high → medium → low)
4. Present actionable items with their IDs

During Implementation

• Reference idea IDs when discussing related tasks
• Cross-check implementation against stored ideas
• Suggest related ideas that could be addressed together
• When a task is complete, remind user to mark it done

After Completion

When user completes a task related to an idea:

1. Identify the matching idea by ID or content
2. Suggest running: is done <id>
3. Confirm the idea was marked as completed

Example Interactions

User: "What should I work on?"

Copy→ Read .ideas.json
→ Filter active ideas
→ Sort by priority
→ Present: "Here are your high-priority ideas: ..."

User: "Check IS for auth ideas"

Copy→ Search ideas containing "auth" or tagged "authentication"
→ Present matching ideas with context

User: "I just finished the login feature"

Copy→ Find related idea about login
→ Suggest: "Should I mark idea abc123 as done? Run: is done abc123"

User: "Add to IS: implement dark mode"

Copy→ Run: is add "implement dark mode" -t ui,feature
→ Confirm idea was added

Installation

Install IDEA Storer CLI:

Copy# From crates.io
cargo install ideastorer

# Or download binary from releases
# https://github.com/byigitt/ideastorer/releases

Tips

• Use short IDs (first 4+ characters) - is done abc1 works if unique
• Tags help organize: -t bug,urgent or -t feature,backend
• Priority levels: low, medium (default), high
• Use --json flag for machine-readable output
• Global storage (--global) for cross-project ideas