Do I need a specification before using this skill?

Yes, this skill requires an existing Dafny specification, often created using the /spec-iterate command or provided directly in the argument.

Does it support languages other than Dafny?

The skill is specifically designed for Dafny, though it includes checks for common pitfalls when extracting Dafny code to target languages like Go or Python.

How does it handle verification errors?

The skill analyzes specific Dafny error types, such as postcondition failures or termination issues, and applies targeted repair strategies like strengthening invariants or adding lemmas.

What does the /generate-verified skill do?

It generates a Dafny implementation that matches a provided specification and iteratively adds proof elements until the Dafny verifier confirms the code is correct.

Dafny Verified Implementation

Name: Dafny Verified Implementation
Author: nicholls-inc

bynicholls-inc

•

安全与测试

Generates and verifies Dafny implementation bodies by iteratively applying proof hints, loop invariants, and lemmas until the verifier accepts.

The generate-verified skill acts as a formal verification expert for developers working with Dafny. It takes an approved specification and writes a corresponding implementation that is mathematically proven to be correct. By iteratively analyzing verifier errors, the skill automatically inserts the necessary loop invariants, assertions, and lemmas required for a successful proof, significantly reducing the manual effort involved in formal methods while providing detailed metrics on proof difficulty and resource usage.

主要功能

013 GitHub stars

02Iterative proof refinement loop (up to 5 attempts)

03Structured verification checklist for production readiness

04Detailed analysis of solver time and resource consumption

05Automatic generation of loop invariants and ghost state

06Target-language compatibility checks for Go and Python extraction

使用场景

01Automating the complex task of finding loop invariants for recursive functions

02Developing mission-critical algorithms that require formal correctness guarantees

03Validating and fixing unverified Dafny code blocks against their specifications

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add nicholls-inc/claude-code-marketplace generate-verified

For use in Claude.ai and ChatGPT

主要功能

013 GitHub stars

02Iterative proof refinement loop (up to 5 attempts)

03Structured verification checklist for production readiness

04Detailed analysis of solver time and resource consumption

05Automatic generation of loop invariants and ghost state

06Target-language compatibility checks for Go and Python extraction

使用场景

01Automating the complex task of finding loop invariants for recursive functions

02Developing mission-critical algorithms that require formal correctness guarantees

03Validating and fixing unverified Dafny code blocks against their specifications

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add nicholls-inc/claude-code-marketplace generate-verified

For use in Claude.ai and ChatGPT