← AI agents

Research tool

PolyClaw: research related contracts through OpenClaw

PolyClaw is a trading-enabled OpenClaw skill for Polymarket. Python and CLI commands combine public research with Polygon wallet operations, a split-and-sell execution path and LLM-assisted searches for logically covering market pairs.

Available on
CLI

Screenshots and interface

What is PolyClaw?

Its buy flow splits collateral into both complementary outcomes and sells the unwanted leg through CLOB. The amount therefore enters a split workflow rather than a simple direct ask-side purchase. Contract approvals also submit on-chain transactions, so the execution mechanics matter during configuration.

Pair analysis and hedge scans use a model to inspect logical implications between precise market questions. Coverage tiers are analytical outputs, not settlement commitments. Entries are tracked in a local JSON file; reconcile wallet activity performed elsewhere before relying on that position list.

Who it is for

  • OpenClaw users seeking Polymarket tools.
  • Developers studying logically related contracts.
  • Technical users configuring RPC and model access.

Features and coverage

  • Market commands

    Search keywords, inspect one market or list trending contracts.

  • Wallet status

    Read collateral and gas-token balances.

  • Split-and-sell buying

    Split collateral and sell the unwanted outcome through CLOB.

  • Local position records

    Store entries locally and display their current values.

  • Pair analysis

    Ask the LLM-backed analyzer about a specified market pair.

  • Hedge scan

    Scan a bounded market set for logical implications and coverage tiers.

  • Early exits

    Sell a held position through available book liquidity.

Workflows with PolyClaw

Research a covering pair

Run a small topic scan, inspect the exact implication and resolution rules, then compare combined cost and size.

Reconcile before exit

Inspect the skill’s entry and current value, then compare local records with the wallet before selling.

Configure services

Use RPC and signer for wallet actions and the model-provider key for hedge analysis.

Pricing and total cost

The source is Apache-2.0 licensed. Polygon transactions, RPC usage, model-provider calls and applicable trading costs are separate.

Getting started

Before you start: The buy command uses split-and-sell mechanics.

  1. Install the skill through the documented OpenClaw route and install Python dependencies with uv.
  2. Configure Chainstack RPC and the optional model-provider key.
  3. Browse before adding execution credentials, then inspect wallet status and contract approvals.
  4. Set trade amounts with split-and-sell mechanics in mind.

Buy amounts enter a split workflow rather than direct ask-side spending.

Useful links

What to check

  • The buy command uses split-and-sell mechanics.
  • Approvals send on-chain transactions.
  • Local JSON positions may omit trades made elsewhere.
  • Coverage tiers are model/logic outputs, without settlement commitments.
  • This skill is for OpenClaw; Alphapoly is a separate standalone application.

A closer look

Ratings & reviews

Editorial rating
3.0/ 5
0 user reviews

Our take on PolyClaw

The inspectable skill links research with position workflows. Split-based buying, local records and LLM pair analysis make it a specialist development tool.

Based on documented features, setup and access terms. How we rate

User reviews 0

The main score switches to user ratings after 5 published reviews.

Be the first to share your experience

A specific example helps the next person choose the right tool.

Related tools

Browse category

Frequently asked questions about PolyClaw

Is a private key needed for search?

Public discovery is separate from wallet execution.

What does buy do?

It splits collateral into both outcomes and sells the unwanted side.

Does the position list include external trades?

It tracks local skill records; reconcile other activity separately.

What uses the model key?

Logical-implication and hedge analysis.