Strict API Verification

SkillAI & models

Your AI stops writing code that calls functions, imports, or variables that do not exist in the version you have installed. This skill keeps the agent limited to methods and imports that provably exist in your setup. You can call on it whenever you want a reality check, for example by saying 'no hallucinations', 'verify APIs', or 'don't invent functions'.

Available today. Use it from your connected AI after setup.

After adding the skill, say something like 'verify APIs' or 'no hallucinations' when you want your AI to stick to what is actually installed. It will then hold itself to the functions, imports, and variables that exist in your version.

Then ask your AI: use the Strict API Verification skill

What your AI can do with it

  • Stop code from calling methods that do not exist in your installed version
  • Leave out imports that are not present in the version you have
  • Avoid variables that cannot be proven to exist in your setup
  • Verify that functions and imports exist in your installed version before using them
  • Respond when you ask for a reality check, such as 'no hallucinations' or 'don't invent functions'

What this skill tells your AI

The instructions your AI receives, as published by alirezarezvani/claude-skills in engineering/strict-api/SKILL.md and read by ahel’s review.

Inventing a function that doesn't exist is the opposite of efficiency. You wrote a line that looks minimal. You shipped a bug that takes an hour to debug. The true minimal path is: use only what is provably there.

Overview

This skill is a reality-check layer applied before any code is written. It is not about being slow — it is about being correct the first time. Use it alongside minimalist when the user wants both less code and verified code.

The Only Rule

Before you write any function call, import, or method access, you must be able to answer:

"Does this exist in the version the user is running?"

If the answer is "probably" or "I think so" — stop. You don't know. Say so.

What This Blocks

Made-up methods:

  • fs.readFileLines() does not exist in Node.js.
  • path.combine() is .NET, not Node.js.
  • csv.read_csv() is pandas, not Python's csv module.

Writing these is not minimal code — it is confident garbage.

Framework confusion. Every framework has a twin that sounds like it:

  • render_template (Flask) vs render() (Django)
  • useForm() (react-hook-form) vs nothing built into React
  • app.listen() (Express) vs server.listen() (raw Node.js http)

Deprecated APIs. Writing a deprecated method is writing code that will break on the next upgrade.

Workflow

  1. Identify every API surface in the code you are about to write: imports, method calls, class instantiations.
  2. Verify each one against the user's stated version. If no version is stated, ask once.
  3. Flag anything uncertain with an inline comment rather than silently guessing.
  4. Prefer verbose-but-correct over terse-but-wrong.

When you are not sure if a method exists, annotate it inline:

// verify fs.openAsBlob exists in your Node.js version (>= 20.0)
const blob = await fs.openAsBlob(path);

One comment costs nothing. A silent wrong call costs an hour of the user's time.

If the uncertainty is too high to write correct code without guessing, say:

"I'd need to check whether X exists in version Y before using it. What version are you on?"

Anti-Patterns

Anti-PatternWhat to do instead
Writing a method call you vaguely rememberStop and verify the exact signature
Silently using a deprecated APIUse the current API and note the deprecation
Assuming API parity across frameworksExplicitly name the framework and version
Guessing import pathsCheck the package's actual export structure
Using an API from a different language's stdlibVerify it exists in this language
Writing "it should work" without checkingAsk what version the user is on

Cross-References

  • Related: engineering/minimalist — use together: minimalist reduces code volume; strict-api ensures what is written is correct.
  • Related: engineering/zero-hallucination-coder — similar goal; broader hallucination prevention beyond APIs.
  • Related: engineering/karpathy-coder — Karpathy-inspired behavioral guardrails for LLM-assisted coding.

Signals

GitHub stars
26k
Forks
4k
Last commit
Aug 2026
Advanced
Catalog kind
skill
Gateway key
strict-api
Source
github.com/alirezarezvani/claude-skills