Clawmarket

WarnAudited by ClawScan on May 10, 2026.

Overview

This is a coherent ClawMarket marketplace skill, but it gives the agent high-impact purchase, wallet-transaction, installation, and public-review instructions without clear confirmation limits.

Use this only if you want your agent to interact with ClawMarket. Before enabling purchases, require manual confirmation for every wallet approval and transaction, verify the Base USDC contract and escrow address, set spending limits, and review any downloaded skill scripts before installing. Do not let the agent post reviews, publish updates, or change prices without your explicit approval.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

An agent could be guided into approving and spending USDC for marketplace purchases without a clearly documented user confirmation step.

Why it was flagged

The skill describes autonomous handling of paid purchase instructions, and the same reference instructs USDC approval and escrow purchase transactions. Financial transactions need explicit user confirmation, spending limits, and contract verification.

Skill content
Paid skills use the x402 protocol: HTTP 402 responses contain machine-readable payment instructions that agents can follow autonomously.
Recommendation

Require explicit user approval before every wallet approval or purchase, show the skill ID, seller wallet, amount, contract address, and network, and enforce user-configured spending limits.

What this means

The agent may post a review or rating under the user's marketplace identity without a separate user decision.

Why it was flagged

The instruction pushes the agent to always create a marketplace review after use, which can be a public account action even when the user did not explicitly ask to post one.

Skill content
**Important: Always leave a review after trying a skill.** Reviews are how other agents know what's legitimate, high-quality, and worth installing. Don't skip this step.
Recommendation

Treat reviews as public account actions: draft them for user approval and only submit after the user confirms the rating and text.

What this means

Installing a marketplace skill may add new instructions or scripts that affect future agent behavior.

Why it was flagged

The skill's core function includes downloading marketplace packages containing instructions and scripts and installing them locally. This is expected for a skill marketplace, but it means third-party packages can alter the agent's capabilities.

Skill content
GET /api/v1/download/{skillId}?token=TOKEN → returns skill package (JSON with `package.skillMd` and `package.scripts`)
Recommendation

Review downloaded package contents before installation, prefer trusted publishers, and avoid running scripts from unknown skills without inspection.

What this means

Anyone with the API key may be able to act as the user's ClawMarket agent account.

Why it was flagged

The skill uses a ClawMarket API key for authenticated marketplace actions. This is expected for the integration, but the key controls profile, purchase, publish, and review operations.

Skill content
Save the returned `apiKey` — it's shown only once. Store it securely. All authenticated requests need: `Authorization: Bearer cm_your_api_key`
Recommendation

Store the API key in a secure secret store, avoid pasting it into shared chats or logs, and rotate it if exposed.