← Back to Skills Marketplace
66
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install screenshots-for-ai-agents
Description
Browser screenshots and vision over MCP Streamable HTTP with x402 USDC micropayments at this deployment. Paid tools require an x402-aware MCP client and a fu...
Usage Guidance
This skill is coherent for remote, paid screenshot capture, but take these precautions before using it:
- Read https://screenshotx402.com/discovery.json and call the free health tool to confirm pricing, network (e.g., base), and mcp.url before any paid call.
- Understand that any URL you submit (and prompts for analyze_screenshot) will be fetched and rendered on the operator's infrastructure — do NOT submit non-public or sensitive URLs or URLs containing secrets.
- Paid features require an x402-aware MCP client and on-chain USDC signing. Confirm how your platform will provide/store signing keys (do not paste private keys into prompts). Prefer a secure platform vault; verify who controls the private keys.
- The skill metadata did not list required environment variables even though the docs reference AGENT_PRIVATE_KEY / X402_PRIVATE_KEY; ask the publisher or your platform how those credentials are expected to be provided and whether any automatic billing/charges can occur.
- Test with the free health endpoint first and verify the deployment identity and TLS cert for https://screenshotx402.com. If you need to capture sensitive content, prefer local/self-hosted tooling instead of this remote paid service.
- If you are uncertain about the facilitator or on-chain flow, consult an ops/devops person familiar with MCP/x402 before enabling paid calls.
Capability Analysis
Type: OpenClaw Skill
Name: screenshots-for-ai-agents
Version: 1.0.1
The skill bundle (SKILL.md) describes a remote screenshot service that requires USDC micropayments via the x402 protocol, instructing the agent to use private keys (e.g., `AGENT_PRIVATE_KEY`) for signing transactions. It directs all data, including URLs and vision prompts, to an external endpoint (https://screenshotx402.com/mcp) for off-host processing. While these behaviors are aligned with the stated service purpose, the requirement for wallet secrets and the potential for unauthorized financial transactions or data exposure via the remote service represent significant security risks.
Capability Tags
Capability Assessment
Purpose & Capability
The skill claims to provide remote browser screenshots via an MCP Streamable HTTP endpoint with x402 USDC micropayments — and its instructions describe exactly that flow, so the requested capabilities match the stated purpose. However, the registry metadata declares no required environment variables or primary credential even though the SKILL.md explains paid tools require a signing wallet (examples: AGENT_PRIVATE_KEY, X402_PRIVATE_KEY). The absence of declared required credentials in metadata is an inconsistency you should confirm with the publisher.
Instruction Scope
Instructions explicitly send submitted URLs and prompts to the remote operator (https://screenshotx402.com) which fetches and renders pages and processes screenshots off-host. The SKILL.md does warn not to submit private or auth-gated URLs, but this off-host processing and transmission of page content and prompts is an inherent privacy risk and should be treated as expected behavior before using the skill.
Install Mechanism
This is an instruction-only skill with no install spec and no code files — nothing is written to disk by the skill itself. That reduces installation risk.
Credentials
While the registry lists no required environment variables, the runtime docs state paid features require an x402-aware MCP client and signing keys (examples: AGENT_PRIVATE_KEY, X402_PRIVATE_KEY) and USDC on-chain. Requesting private keys and on-chain funds for paid capture is proportionate to the paid feature, but the metadata omission means the skill may prompt you to supply sensitive credentials or platform vault entries at runtime; verify how your platform will store/require those secrets.
Persistence & Privilege
The skill does not request permanent/always-on inclusion (always: false). It does not request to modify other skills or system-wide settings. Autonomous invocation is allowed (platform default); combine that with paid operations only if you trust the deployment.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install screenshots-for-ai-agents - After installation, invoke the skill by name or use
/screenshots-for-ai-agents - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
**Security, privacy, and usage requirements clarified for paid tools.**
- Added explicit guidance about privacy: only use for public, non-sensitive URLs, as all page data and prompts are processed off-host.
- Stated requirement for x402-aware MCP client and funded wallet: paid endpoints require signing USDC payments via agent environment, never in user prompts.
- Warned users not to store keys or secrets in prompts—use secure environment variables as per documentation.
- Clarified that free discovery and health checks require no wallet or signing access.
- Updated description and onboarding steps to emphasize security and payment prerequisites.
v1.0.0
Initial release of screenshot-x402-mcp skill.
- Provides browser screenshots and vision analysis via MCP Streamable HTTP with x402 USDC micropayments.
- Supports three tools: free health check, paid screenshot capture (PNG/JPEG), and paid screenshot analysis (image + vision text).
- Integrates automatic x402 payment flow for tool usage without manual payment header handling.
- Discovery endpoints and API documentation available at https://screenshotx402.com.
- Requires public HTTPS URLs for capture; vision output is server-configured.
Metadata
Frequently Asked Questions
What is Screenshots for AI agents via x402?
Browser screenshots and vision over MCP Streamable HTTP with x402 USDC micropayments at this deployment. Paid tools require an x402-aware MCP client and a fu... It is an AI Agent Skill for Claude Code / OpenClaw, with 66 downloads so far.
How do I install Screenshots for AI agents via x402?
Run "/install screenshots-for-ai-agents" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Screenshots for AI agents via x402 free?
Yes, Screenshots for AI agents via x402 is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Screenshots for AI agents via x402 support?
Screenshots for AI agents via x402 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Screenshots for AI agents via x402?
It is built and maintained by buildinhk (@buildinhk); the current version is v1.0.1.
More Skills