MasterCard | Is your claw a shopaholic?

Let your agent shop online with guardrailed wallets, multiple payment methods, and owner approval.

Key things to check before installing: - Confirm the publisher and domain: the skill's files point to creditclaw.com but the registry name/slug mention 'MasterCard' — verify this mismatch with the publisher or registry metadata. - Treat downloaded files as code: the flow requires saving files and executing a decrypt.js delivered with encrypted card files. Only run such scripts in a sandboxed environment and inspect the script contents before execution. - Prefer the sub-agent flow: the docs recommend spawning an ephemeral sub-agent so the main agent never sees decrypted card details. If your agent platform cannot isolate sub-agents, avoid enabling this skill or avoid using rail5/encrypted-card functionality. - Limit the API key: give the skill a dedicated CREDITCLAW_API_KEY with least privilege and monitor usage. Do not reuse a high-privilege or multi-service key. - Review guardrails: ensure approval_mode and per-transaction/daily limits are strict (ask_for_everything or low thresholds) before permitting autonomous purchases. - If you are unsure about the domain or publisher identity, do not install. Request clarification from the registry owner/maintainer or use a vetted alternative.

Type: OpenClaw Skill Name: mastercard Version: 1.0.2 The bundle describes a financial enablement platform for AI agents to manage spending and payments via encrypted cards and USDC. It features robust security protocols, such as server-side enforcement of spending limits, single-use decryption keys, and the use of ephemeral sub-agents to isolate sensitive information. The instructions in SKILL.md and ENCRYPTED-CARD.md are transparent about security risks and provide clear guidance on protecting API keys, with no evidence of malicious intent or unauthorized data exfiltration (IOC: creditclaw.com).