Generate 10 Ideas

Process

1. Understand the system — Read available documentation, code structure, and existing capabilities to deeply understand what the system does today, who uses it, and what its constraints are.
2. Generate 100 candidates — In your extended thinking, brainstorm ONE HUNDRED distinct ideas across these dimensions:
   • User experience and workflow improvements
   • New capabilities and integrations
   • Performance, reliability, and robustness
   • Developer experience and extensibility
   • Automation and intelligence
   • Composability and interoperability
   • Error handling and recovery
   • Observability and debugging
   • Security and safety
   • Novel interaction patterns
   For each idea, briefly assess: impact (high/medium/low), implementation difficulty (easy/medium/hard), and complexity burden introduced.
3. Filter ruthlessly — From the 100 candidates, select exactly 10 that score highest on this combined criteria:
   • Brilliance: genuinely clever or non-obvious insight
   • Impact: meaningfully improves the system for users
   • Pragmatism: achievable without extreme effort or fragile complexity
   • Innovation: not just an incremental tweak — a real leap in capability or usability
4. Present the top 10 — For each idea, provide:
   • Title: short, memorable name
   • One-liner: what it does in one sentence
   • Why it's brilliant: the non-obvious insight that makes this powerful
   • Implementation sketch: 2-3 sentences on how you'd build it
   • Complexity cost: honest assessment of what this adds to the system

Output Format

### #N — [Title]

**What:** [one-liner description]

**Why it's brilliant:** [the insight]

**How to build it:** [implementation sketch]

**Complexity cost:** [honest assessment]

Constraints

• Every idea must be implementable by a small team in days-to-weeks, not months.
• No ideas that require fundamental architectural rewrites.
• No ideas that add complexity without proportional value.
• Prefer ideas that compose well with existing functionality.
• Prefer ideas that unlock further possibilities (force multipliers over dead ends).
• Be specific to the actual system you're looking at — generic advice like "add caching" or "improve error messages" is not acceptable unless the specific application is genuinely novel.