← Back to Skills Marketplace
realfishsam

Molt Pmxt

by realfishsam · GitHub ↗ · v1.1.0
cross-platform ⚠ suspicious
1273
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install molt-pmxt
Description
Grants the agent real-time access to prediction markets (Polymarket, Kalshi, Limitless) for fact-checking, probability analysis, and order execution.
Usage Guidance
Before installing or enabling this skill: 1) Verify the author/source — the registry metadata lacks an owner homepage and does not declare the sensitive environment variables that SKILL.md and the code use. 2) Do not provide high-value private keys — if you must test, use a separate test wallet with minimal funds or read-only / limited-scope API credentials. 3) Ask the maintainer to explain why Kalshi needs an RSA private key and to correct the registry metadata to list required env vars. 4) Review the pmxtjs dependency (maintainer, recent versions) and ensure your platform runs dependency installation in a controlled environment. 5) Confirm how and where your credentials will be stored by the agent (in-memory vs persisted), and rotate any keys used for testing. 6) If you lack trust in the publisher, run the package in an isolated sandbox or refrain from enabling order-execution capabilities.
Capability Analysis
Type: OpenClaw Skill Name: molt-pmxt Version: 1.1.0 The skill is designed to interact with real-money prediction markets, requiring sensitive credentials (private keys, API keys) via environment variables and enabling financial transactions. While these are inherently high-risk capabilities, the `SKILL.md` explicitly declares these requirements and, crucially, includes a safety instruction for the AI agent to seek explicit user confirmation before executing any `pmxt_order` (trade). The code (`src/tools.ts`) adheres to the stated purpose, making network calls only to the approved prediction market domains (polymarket.com, kalshi.com, limitless.exchange) and processing user input within the bounds of API calls, without evidence of data exfiltration, malicious execution, or prompt injection for unauthorized actions.
Capability Assessment
Purpose & Capability
The skill claims trading and order execution on Polymarket, Kalshi, and Limitless and the SKILL.md and src/tools.ts clearly expect API keys and private keys for these exchanges. However, the registry metadata lists no required environment variables or primary credential — an internal inconsistency that can hide the true privileges the skill needs.
Instruction Scope
SKILL.md instructs the agent to read several sensitive environment variables (private keys, API keys) and to perform network calls to the exchanges (domains are listed). It does not instruct reading unrelated system files. There is a small contradiction in the guidance: it says to "silently check" arbitrage spreads but also says you MUST alert the user when you detect an arbitrage — this should be clarified. Otherwise the runtime instructions generally stay within the stated purpose.
Install Mechanism
No install specification is provided (lower disk-write risk), but the package.json/package-lock are included and declare a dependency on 'pmxtjs'. That dependency is appropriate for the purpose, but users should confirm whether the platform will run npm install and where node_modules would be written. No suspicious external download URLs or extract steps are present.
Credentials
SKILL.md requires multiple sensitive credentials (POLYMARKET_PRIVATE_KEY, KALSHI_API_KEY, KALSHI_PRIVATE_KEY (RSA), LIMITLESS_API_KEY, LIMITLESS_PRIVATE_KEY) which are proportionate for an exchange-trading skill — but the registry metadata omitted these entirely. Also the requirement of an "RSA private key" for Kalshi is unusual and should be justified. The mismatch between declared metadata and actual env usage increases risk of unnoticed credential exposure.
Persistence & Privilege
The skill does not request always:true and does not appear to modify other skills or system-wide settings. It initializes exchange clients at module load but does not request permanent platform presence beyond normal skill operation.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install molt-pmxt
  3. After installation, invoke the skill by name or use /molt-pmxt
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.1.0
- Updated skill name from "pmxt" to "molt-pmxt" - Changed author to "realfishsam" - Bumped version from 1.0.0 to 1.1.0 - No changes to functionality or instructions; documentation and metadata updates only
v1.0.0
Initial release: real-time prediction market access, probability analysis, and order execution. - Connects to Polymarket, Kalshi, and Limitless for market search, pricing, and trades. - Supports three main functions: market discovery (`pmxt_search`), odds retrieval (`pmxt_quote`), and real-money order placement (`pmxt_order`). - Requires API keys and private keys for trading on supported platforms. - Emphasizes accuracy, clear percentage displays, and checks for arbitrage opportunities automatically. - Includes agent behavioral guidelines for reliable and user-friendly operation.
Metadata
Slug molt-pmxt
Version 1.1.0
License
All-time Installs 0
Active Installs 0
Total Versions 2
Frequently Asked Questions

What is Molt Pmxt?

Grants the agent real-time access to prediction markets (Polymarket, Kalshi, Limitless) for fact-checking, probability analysis, and order execution. It is an AI Agent Skill for Claude Code / OpenClaw, with 1273 downloads so far.

How do I install Molt Pmxt?

Run "/install molt-pmxt" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Molt Pmxt free?

Yes, Molt Pmxt is completely free (open-source). You can download, install and use it at no cost.

Which platforms does Molt Pmxt support?

Molt Pmxt is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Molt Pmxt?

It is built and maintained by realfishsam (@realfishsam); the current version is v1.1.0.

💬 Comments