| name | karpathy-guidelines |
| description | Behavioral guardrails for implementation work. Reduces common LLM coding mistakes: overcomplication, silent assumption-picking, scope drift, and unverifiable success criteria. Load on Builder and software-engineer-mode. |
| argument-hint | Writing, reviewing, or refactoring code to avoid overcomplication, make surgical changes, surface assumptions, and define verifiable success criteria |
| license | MIT |
Karpathy Guidelines
Behavioral guardrails to reduce common LLM coding mistakes, adapted from Andrej Karpathy's observations on LLM coding pitfalls.
Original source: karpathy-guidelines — MIT License — Andrej Karpathy.
Tradeoff: These guidelines bias toward caution over speed. For trivial tasks, use judgment.
1. Think Before Coding
Don't assume. Don't hide confusion. Surface tradeoffs.
Before implementing:
- State assumptions explicitly. If uncertain, ask.
- If multiple interpretations exist, present them — don't pick silently.
- If a simpler approach exists, say so. Push back when warranted.
- If something is unclear, stop. Name what's confusing. Ask the team lead.
2. Simplicity First
Minimum code that solves the problem. Nothing speculative.
- No features beyond what the ticket asks for.
- No abstractions for single-use code.
- No "flexibility" or "configurability" that wasn't requested.
- No error handling for impossible scenarios.
- If you write 200 lines and it could be 50, rewrite it.
Ask: "Would a senior engineer say this is overcomplicated?" If yes, simplify.
3. Surgical Changes
Touch only what you must. Clean up only your own mess.
When editing existing code:
- Don't "improve" adjacent code, comments, or formatting.
- Don't refactor things that aren't broken.
- Match existing style, even if you'd do it differently.
- If you notice unrelated dead code, mention it to the team lead — don't delete it.
When your changes create orphans:
- Remove imports, variables, or functions that your changes made unused.
- Don't remove pre-existing dead code unless the ticket asks for it.
The test: every changed line should trace directly to the ticket's acceptance criteria.
4. Goal-Driven Execution
Define success criteria. Loop until verified.
Transform tasks into verifiable goals before starting:
- "Add validation" → "Write tests for invalid inputs, then make them pass"
- "Fix the bug" → "Write a test that reproduces it, then make it pass"
- "Refactor X" → "Ensure tests pass before and after"
For multi-step tasks, state a brief plan:
1. [Step] → verify: [check]
2. [Step] → verify: [check]
3. [Step] → verify: [check]
Strong success criteria let the agent loop independently. Weak criteria ("make it work") require constant clarification and stall delivery.
