Monzo

**Quick judgement:** This is a well-designed, feature-rich Monzo banking skill for OpenClaw with strong documentation and reasonable security practices. It’s suitable for serious personal finance queries and automations where you explicitly want your agent to operate on a real Monzo account. **Key strengths:** - Broad practical coverage: balances, transactions, pots, feed notifications, receipts, and webhooks. - Very clear setup and usage docs, including examples tailored to agent-style queries. - Attention to security: encrypted credentials, file-permission guidance, HTTPS-only, and reliance on Monzo’s SCA and app approvals. **Main risks & cautions:** - High-impact domain: the skill can move real money between pots and send live notifications. Misuse (e.g., from bad prompts, overbroad instructions, or compromised agent configs) could have financial consequences. - Configuration sensitivity: relies on correct MONZO_KEYRING_PASSWORD handling and secure storage of OAuth client secrets; misconfigured permissions or plaintext configs can weaken security. - Operational friction: SCA approvals, expired/invalid tokens, or insufficient permissions can break flows and require user intervention. **Recommended scenarios:** - You want your OpenClaw agent to answer “live” questions about Monzo balances, recent spending, and category-level activity. - You’re comfortable with OAuth, environment variables, and securing secrets on your OpenClaw host. - You want automations that adjust pots (e.g., moving money into a savings pot after certain triggers) or send Monzo app reminders/alerts as part of larger workflows. **Less suitable when:** - You cannot confidently secure environment variables and config files on the host. - You only need generic budgeting analysis and are fine with non-banking data sources (in which case a non-bank-linked skill is safer).