How do I install the Karpathy Recipe skill?

Run npx skillmds@latest add andersonlimahw/karpathy-recipe in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Karpathy Recipe safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: makes network calls. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Karpathy Recipe?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Karpathy Recipe free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Karpathy Recipe?

Andersonlimahw (@andersonlimahw) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Karpathy Recipe by andersonlimahw — Agent Skill

Name: Karpathy Recipe
Rating: 4.666666666666667 (3 reviews)
Author: Andersonlimahw

Karpathy Recipe

Applies Karpathy's "A Recipe for Training Neural Networks" adapted to software engineering. Forces a minimum runnable baseline before optimizing, one knob at a time, with verifiable eval at the beginning. Use when implementing a new feature, doing a non-trivial refactor, integrating an external service, or when the user says "implement X from scratch", "how to start feature Y", "recipe", "incremental approach", "baseline first".

File contents

karpathy-recipe

Translation of Recipe for Training Neural Networks to product features. Principle: be paranoid, go slow, visualize everything.

Mandatory Flow

1. Become one with the data

Before coding, read real data from the domain:

For a financial feature: 10 real transactions from extratos_exemplo/ or from financial_transactions in staging
For AI: 5 examples of expected input/output from the user
List known edge cases (memory, ADRs, past issues) — do not invent them

Output: 1 paragraph in the PR ## Data I looked at with links/IDs.

2. Set up end-to-end skeleton + dumb baseline

Implement the complete path with trivial logic first:

Server action returns a fixed mock
Repository returns []
UI uses hardcoded data

Validate that the entire pipeline runs (request → action → repo → DB → UI) before any real logic. Do not optimize anything yet.

3. Overfit the simplest case

Next: make it work on 1 real input, ignoring edge cases. Mock externals (Stripe, Belvo, OpenAI). Confirm expected logs/output.

4. Add one knob at a time

Add features in order (validation → real DB → external API → caching → optimization). After each knob:

run rtk npm test
run rtk tsc --noEmit
atomic commit (semantic) — each commit must pass green

If something breaks: revert the knob, do not stack fixes.

5. Define eval up front

Before merging, write:

1 unit test (happy path)
1 integration test (with real fixtures when possible — see memory: "integration tests must hit a real database")
## Eval criteria in the feature doc: "done when X is measurable"

6. Regularize / harden

Only now add: rate limit, error boundary, retry, Sentry observability, i18n keys, extra RLS. Not before.

Forbidden Anti-patterns

Starting optimization before the baseline is running
Adding two knobs in the same commit
"I will add a try/catch just in case" without a reproduced error
Caching before having performance measurements
Generic abstractions (factory, strategy) without 3 real callsites

Expected Output when this skill is active

Codex responds with a mini-plan:

Recipe plan:
1. Data: <source of 5-10 real examples>
2. Skeleton: <files to create with mocks>
3. Overfit: <1 case to make work>
4. Ordered knobs: [...]
5. Eval: <test + measurable criteria>

And asks for confirmation before coding.

[[surgical-reviewer]] — reviewer that validates these principles in a PR
[[eval-harness-builder]] — agent that generates tests for step 5
[[andrej-karpathy-skills:karpathy-guidelines]] — base behavioral guidelines

---name: karpathy-recipedescription: Applies Karpathy's "A Recipe for Training Neural Networks" adapted to software engineering. Forces a minimum runnable baseline before optimizing, one knob at a time, with verifiable eval at the beginning. Use when implementing a new feature, doing a non-trivial refactor, integrating an external service, or when the user says "implement X from scratch", "how to start feature Y", "recipe", "incremental approach", "baseline first".---# karpathy-recipeTranslation of [Recipe for Training Neural Networks](https://karpathy.github.io/2019/04/25/recipe/) to product features. Principle: **be paranoid, go slow, visualize everything**.## Mandatory Flow### 1. Become one with the dataBefore coding, read real data from the domain:- For a financial feature: 10 real transactions from `extratos_exemplo/` or from `financial_transactions` in staging- For AI: 5 examples of expected input/output from the user- List known edge cases (memory, ADRs, past issues) — do not invent themOutput: 1 paragraph in the PR `## Data I looked at` with links/IDs.### 2. Set up end-to-end skeleton + dumb baselineImplement the **complete path** with trivial logic first:- Server action returns a fixed mock- Repository returns `[]`- UI uses hardcoded dataValidate that the entire pipeline runs (request → action → repo → DB → UI) before any real logic. **Do not optimize anything yet.**### 3. Overfit the simplest caseNext: make it work on **1 real input**, ignoring edge cases. Mock externals (Stripe, Belvo, OpenAI). Confirm expected logs/output.### 4. Add one knob at a timeAdd features in order (validation → real DB → external API → caching → optimization). After each knob:- run `rtk npm test`- run `rtk tsc --noEmit`- atomic commit (semantic) — each commit must pass greenIf something breaks: **revert the knob**, do not stack fixes.### 5. Define eval up frontBefore merging, write:- 1 unit test (happy path)- 1 integration test (with real fixtures when possible — see memory: "integration tests must hit a real database")- `## Eval` criteria in the feature doc: "done when X is measurable"### 6. Regularize / harden**Only now** add: rate limit, error boundary, retry, Sentry observability, i18n keys, extra RLS. Not before.## Forbidden Anti-patterns- Starting optimization before the baseline is running- Adding two knobs in the same commit- "I will add a try/catch just in case" without a reproduced error- Caching before having performance measurements- Generic abstractions (factory, strategy) without 3 real callsites## Expected Output when this skill is activeCodex responds with a mini-plan:```Recipe plan:1. Data: <source of 5-10 real examples>2. Skeleton: <files to create with mocks>3. Overfit: <1 case to make work>4. Ordered knobs: [...]5. Eval: <test + measurable criteria>```And asks for confirmation before coding.## Related- [[surgical-reviewer]] — reviewer that validates these principles in a PR- [[eval-harness-builder]] — agent that generates tests for step 5- [[andrej-karpathy-skills:karpathy-guidelines]] — base behavioral guidelines

Frequently asked questions