sui: Skill for Claude Code

.claude/skills/dual-replay/SKILL.md

dual-replay is a skill for Claude Code from MystenLabs/sui. It costs 27 tokens per session (1,589 once invoked), scanned A, original, Apache-2.0.

A test tool that runs Sui transactions through two versions of the execution system and compares their results. Sui is a blockchain platform, and a commit is a saved point in a code repository.

In plain words
What is it for?
Use it to compare a base commit with a newer commit, investigate failed replay steps, and save differences and execution timings.
Why use it?
It reveals whether a code change alters transaction outcomes, status, or gas usage, helping detect unintended behavior before release.

Skill for Claude Code

Written for Claude Code: installed under .claude/.

This is MystenLabs/sui's own configuration. It tells Claude Code how to work on sui itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything sui configures →

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is python3 scripts/dual_replay.py run \.

About the project

Sui is a blockchain platform for running smart contracts and managing digital assets, using the Move programming language and a network of authorities to process transactions. It is intended for developers building web3 applications and on-chain assets.

MystenLabs/sui · 7,744 stars · on GitHub · sui.io

Reuse

Borrowing it

Nothing to install: this file belongs to MystenLabs/sui. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/MystenLabs/sui/main/.claude/skills/dual-replay/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/MystenLabs/sui

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for dual-replay

README.md
[![agentmods](https://agentmods.dev/badge/skills/mystenlabs/sui/dual-replay/github.svg)](https://agentmods.dev/skills/mystenlabs/sui/dual-replay)
Your own site
<a href="https://agentmods.dev/skills/mystenlabs/sui/dual-replay"><img src="https://agentmods.dev/badge/skills/mystenlabs/sui/dual-replay/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for dual-replay

Your own site · 80×15
<a href="https://agentmods.dev/skills/mystenlabs/sui/dual-replay"><img src="https://agentmods.dev/badge/skills/mystenlabs/sui/dual-replay.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 27 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,589 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector warn 7 Sept 2026
SkillSpector: 1 finding, up to medium

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • medium Rogue Agent · line 52
    Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.
    Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00027 $0.01589
Opus 5 $0.00014 $0.00794
Sonnet 5 $0.00005 $0.00318
Haiku 4.5 $0.00003 $0.00159

Measured 11d ago against content hash 37b80ed170a8, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

dual-replay scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/dual-replay/SKILL.md · 96 lines

How it starts

The opening of the file, as written. The whole thing — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Dual Execution Replay

Instruments the latest execution path to run every transaction twice — once with a base executor (built from an execution-layer cut taken at <base-sha>) and once with the tip executor at <tip-sha> — and diffs the two outputs. The base result is committed; the tip result is for comparison only.

Most of the work is handled by scripts/dual_replay/. This skill is the contract for invoking that script and the playbook for when its happy path fails.

Usage

/dual-replay <base-sha> <tip-sha> [diff-mode] [cut-name]
  • diff-mode (optional, default strict): one of strict, status-only, gas-only, status-and-gas.
  • cut-name (optional, default replay_cut).

Additional knobs (pass through if the user supplied them):

  • --gas-tolerance <percent>: tolerated gas delta as a percentage (default 0, i.e. exact equality). Only meaningful for diff modes that compare gas.
  • --output-dir <path>: where effects-diff artifacts go (default target/dual-replay/).
  • --timings-file <path>: where per-tx execution timings are appended as CSV (default <output-dir>/timings.csv).

Timings output

Every transaction's base and tip execution is timed and appended as one CSV row to --timings-file (default <output-dir>/timings.csv). Columns: digest,base_ns,tip_ns,base_gas,tip_gas,status_match. Writes are buffered and flushed every 500 rows. The file is opened in append mode, so resumed runs accumulate into the same file — delete it between runs if a clean dataset is needed.

If either SHA is missing or does not resolve to a commit, ask the user for clarification. Do not prompt for other arguments — fall back to defaults.

Happy path

Run the script's run subcommand. It does everything end-to-end and creates a commit on the current branch:

python3 scripts/dual_replay.py run \
  --base <base-sha> --tip <tip-sha> \
  --cut-name <cut-name> \
  --diff-mode <diff-mode> \
  [--gas-tolerance <percent>] \
  [--output-dir <path>]

Read the full file on GitHub · 96 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 11d ago First seen · 96 lines · 27 tokens per session scan A 37b80ed170a8

Subscribe to this mod's changes

dual-replay is a skill published in the GitHub repository MystenLabs/sui (7,744 stars, last pushed today), licensed Apache-2.0. It adds 27 tokens to every session and 1,589 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

blockchain-application

Use this skill when asked about smart contract development, Solidity, Vyper, Rust smart contracts (Solana, NEAR, Polkadot), Haskell/Plutus (Cardano), Cairo/StarkNet, dApp backend development, Truffle, Hardhat, Foundry, Anchor, and blockchain application patterns. Languages: Solidity, Vyper, Rust, Haskell, Cairo, Move.…

j4flmao/agent-skills · 175 tokens

smart-contract-testing

Guide for testing smart contracts using Foundry and Hardhat, covering unit tests, fuzz testing, invariant testing, fork testing, and gas benchmarking.

nirholas/three.ws · 33 tokens

solidity-playground-guide

Guide to Lyra Web3 Playground — a browser-based Solidity IDE and smart contract playground. Write, compile, deploy, and interact with contracts directly in the browser. Supports Hardhat/Foundry compilation, multiple chains, and live contract testing.

nirholas/three.ws · 55 tokens

blockchain-expert

Expert-level blockchain, Web3, smart contracts, DeFi, and cryptocurrency development. Use when the user mentions Web3, smart contracts, DeFi, Ethereum, or Solidity, or when the task involves Blockchain Fundamentals, Web3 & DeFi, Smart Contract Security, or Gas Optimization.

personamanagmentlayer/pcl · 63 tokens

lineth-quickstart

Operating manual for the Lineth Stack quickstart — the Docker-Compose dev/demo stack at docs/getting-started/lineth-stack in the lineth-monorepo that boots a local Linea/Lineth L2 with Sepolia or local L1 finality. Use whenever you are working inside the lineth-stack quickstart and need to boot or run the stack…

LFDT-Lineth/lineth-monorepo · 228 tokens

blockchain-testing

Use this skill when asked about testing smart contracts, Foundry tests, Hardhat tests, fuzz testing, invariant testing, property-based testing, formal verification, audit preparation, integration testing for dApps, and blockchain testing patterns. Covers Foundry cheatcodes, Echidna fuzzing, Certora verification…

j4flmao/agent-skills · 117 tokens