Piggy

@Piggy what's my current balance and how much have I spent this month across all categories?

Based on your Piggy account data, your current balance is $2,847.32. This month you've spent $1,204.56 across all categories: $412 on groceries, $298 on dining out, $187 on transportation, $156 on utilities, and $151.56 on miscellaneous expenses. Your spending is tracking 8% below your monthly average.

This read-heavy example shows balance and spending aggregation. Requires API key with read access to transaction history. Output reflects account state at query time — balances and totals will differ by user and date range.

@Piggy create a new savings goal called 'Emergency Fund' with a target of $5000 and set aside $200 from my checking account as the first deposit.

I've created a new savings goal 'Emergency Fund' in your Piggy account with a $5,000 target. The initial $200 deposit from your checking account has been recorded, bringing your goal progress to 4%. You'll receive monthly reminders to review progress toward this goal.

This write example demonstrates goal creation and fund allocation. Requires API key with write permissions. Side-effect: modifies account state and may trigger notification preferences. Confirm the deposit amount aligns with available balance before executing.

@Piggy analyze my spending patterns over the last three months and suggest two areas where I could realistically cut back by 10% without major lifestyle changes.

After reviewing your last three months of transactions, two opportunities stand out: dining out ($894 monthly average) and subscription services ($127 monthly average). Reducing dining out by 10% would save $89/month — achievable by cooking one extra meal per week. Cutting subscriptions by 10% ($13/month) could mean pausing one service you rarely use. Combined, these changes would free up $102/month with minimal disruption.

This synthesis example pairs transaction data with AI reasoning to surface actionable insights. Effectiveness depends on transaction categorization accuracy in Piggy. Suggestions are illustrative — actual recommendations vary by spending history and user context.

When you need a placeholder before real integration work starts

A 3-person startup is mapping out their product roadmap and wants to reference Piggy's capabilities in planning docs before they commit engineering time. This MCP is useful if you're in discovery mode—your team can ask questions about Piggy's API surface, check what endpoints exist, and sketch workflows in Switchy without writing code. The trade-off: with zero tools exposed right now, you're limited to whatever the MCP's underlying model knows about Piggy's public docs. If your team needs to actually call Piggy endpoints or pull live data, you'll hit a wall until tools get added. Use this MCP when you're researching vendor fit, not when you're shipping features that depend on Piggy.

Comparing Piggy against alternatives in a shared workspace

A 6-person finance team is evaluating three payment or loyalty platforms and wants a single place to compare API capabilities. They add the Piggy MCP to a Switchy workspace alongside MCPs for competitors, then ask side-by-side questions about rate limits, webhook schemas, and auth flows. This works because the team isn't building yet—they're just gathering facts to inform a build-or-buy decision. The API key requirement means someone on the team already has a Piggy account, which is fine for eval but assumes you're past the cold-outreach stage. If your team is still in the 'should we even talk to this vendor' phase, the MCP won't add much until you have credentials and a clearer sense of what Piggy does.

When support needs to explain Piggy integrations to clients

A 10-person SaaS company uses Piggy in their product, and their support team fields questions about how the integration works. They add the Piggy MCP to a Switchy workspace so support reps can quickly pull integration details without pinging engineering. This scenario assumes Piggy's docs are well-structured and the MCP can surface them—but with no tools live yet, you're relying on the model's training data, which might be stale or incomplete. If your support volume is low (under 5 Piggy questions a week), this is overkill; just bookmark the docs. If it's higher and you need faster answers, the MCP is worth trying, but expect to add internal notes as tools get built out.