Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
ClawTasks is an Agent-to-Agent Bounty Marketplace operated by AI Magic, LLC. Its goal is to let AI Agents post, claim, and complete tasks via API. It was originally built around on-chain escrow and settlement in USDC on Base L2, but the site repeatedly notes that it has now switched to free-task only mode. Paid bounties are being wound down, with the focus shifting to improving reliability, review workflows, and worker quality.
The platform supports both standard bounties and Metric Bounties. The former are suited to deliverable-based tasks such as coding, content, research, and design, while the latter target measurable outcomes such as views, sign-ups, and engagement rates. Claiming modes include Instant Claim and Proposal Mode. The full workflow involves the poster depositing USDC on-chain, the Worker claiming the task and staking 10%, submitting a text-based result of up to 50,000 characters, and the poster reviewing it. If the poster does not respond within 48 hours, the submission is automatically approved. The API documentation includes endpoints for registering Agents, listing tasks, creating tasks, claiming tasks, confirming stakes, submitting work, and approving results. This makes it more of a machine-callable developer tool than a traditional human crowdsourcing platform.
The current product announcement clearly states that it operates in free-task mode. Under the historical paid model, Workers received 95% of the bounty and got their stake back after completing a task, while the platform charged a 5% fee. Rejections, timeouts, or abandonment could result in loss of the stake. The terms emphasize that this is beta experimental software, with risks including AI misoperation, smart contract vulnerabilities, irreversible on-chain transactions, and third-party service failures. Users must supervise their own Agents and bear the risk of financial loss.
Its strengths are its forward-looking positioning, clear API workflow, and relatively complete design around on-chain escrow, staking, and automatic approval, making it suitable for exploring Agent-based economic collaboration. The downsides are also obvious: paid functionality is currently being scaled back, reliability is still under development, and the reliance on wallets, USDC, Base, and gas creates a high barrier to entry. File submissions depend on external links, and there does not appear to be a complete set of engineering documentation covering SDKs, authentication details, error codes, and similar essentials.
ClawTasks is best suited to AI Agent developers, autonomous-task experimentation teams, and Web3 developers running small-scale trials. It is not ideal for enterprises that require stable SLAs, fiat settlement, or mature customer support. The source material does not provide details on access from China. Since it involves cryptocurrency, the Base network, and USDC, network accessibility, compliance, and payment availability all need to be verified independently. If you only need task collaboration or bounty management, alternatives to compare include Gitcoin, Dework, Questbook, Bountycaster, or GitHub issue bounty workflows.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on clawtasks.com official site.
clawtasks.com is an United States AI Apps provider. TG4G tracks its product information, an overall rating of 6.0/10, and a China-accessibility score of Workable. Click "Visit Official Site" to reach clawtasks.com directly.