SwitchWize MCP server
online · 100% uptimeio.github.switchwize/switchwize-mcp·v0.3.0·1 starsSixteen read-only US rate, savings-gap, Fed, card and CD tools with sources and freshness.
1 stars54 downloads/wk
Reviews
Write oneNobody has reviewed SwitchWize MCP yet.
If you have run it, two minutes of your experience saves the next person an afternoon.
SwitchWize MCP tools (16, 1 write)
write = sends, deletes, buys or postsbest_move_this_weekFreeGiven the credit cards a user holds, rank this week's dollar-valued actions: which held card to use per spend category, welcome offers at a 12-month high on cards not held, active transfer bonuses, and (optionally) the savings-rate move beside the card move.
compare_switch_savingsFreeCompare a user's current APY with SwitchWize's current top savings APY for a balance.
get_bank_gapFreeEstimate annual savings interest forgone for a supplied balance and optional bank name. comparisonApy is the average of the top three tracked savings rates, not an average across all banks; bankCount is the size of the tracked set. baselineApy is the named bank's APY, or the national savings benchmark when no bank is supplied. Return annualGap, APYs, observation date, freshness and methodology. This is a hypothetical estimate, not guaranteed savings or an account recommendation; verify eligibility, fees and current provider terms.
get_bank_gap_indexFreeReturn the current SwitchWize Bank Gap Index headline value, latest release, and methodology URL.
get_card_changesFreeReturn recent verified credit-card fee and APR changes, most recent first. Each event is a diff between two confirmed snapshots of the same card.
get_card_downgrade_pathFreeReturn verified credit-card product-change (downgrade) paths and the issuer's fee-refund window: which no-fee or lower-fee card a given card can be product-changed to, and how long after the annual fee posts a refund is available on cancellation. Optionally filtered to one card id (issuer:slug); returns every verified path and window when omitted.
get_card_offersFreeReturn current credit-card welcome-bonus offers (non-targeted, confidence-gated), each with a verified timestamp, source URL, and the card's own 12-month best.
get_cd_rollover_rulesFreeReturn verified CD rollover rules for one bank, or every bank SwitchWize has verified: what happens automatically at maturity if the customer does nothing, the grace period, how to stop the rollover, and the early-withdrawal penalty. Each fact is sourced directly from the bank's own disclosure; an unverified field is omitted, never guessed.
get_fed_hike_pass_throughFreeReturn how much of the latest Federal Reserve rate hike each tracked savings account has passed through to savers so far: rate before the hike, rate now, change in basis points, pass-through share, and days until the bank moved. One source per bank; each row carries its source type and observation dates. Optionally filtered to one institution id.
get_index_datasetFreeReturn a SwitchWize index dataset summary and recent series.
get_rate_freshnessFreeReturn public data source freshness status and affected routes.
get_reality_scoreFreeEstimate the modeled annual value of a supported premium credit card for a consumer segment. Supply a card slug or name and optionally a segment ID; the default is occasional-traveler. Return segment assumptions, score, rewards, credits, perks, fee, interest, friction, net value, caveats, sources, verification date, freshness and methodology. Gross recovered value is rewards + credits + perks. breakageMultiplier is gross recovered value / advertised value, NOT divided by annual fee. feeRecoveryMultiplier is max(0, gross recovered value - interest drag - rewards friction discount) / annual fee, before subtracting the fee. Friction applies to raw rewards only. Reality Score is 100 * net realized value / advertised value. Ratios are rounded. These are scenario estimates, not personalized advice or guaranteed benefits. Preserve stale or unavailable labels; do not infer that all terms share the observation date. Unknown cards or segments return a structured error with suggestions.
get_reward_valuationsFreeReturn SwitchWize points/miles valuations (cents per point) per reward currency, with sample size and provisional/published status.
get_top_hysa_ratesFreeReturn top high-yield savings rates with freshness and attribution. Answers only from the editor-approved snapshot; before its first release it returns no ranked rates and says so (approval_status: not_yet_approved). approvedOnly makes the requirement explicit. detail=full adds every listed account marked verified or tracked, best picks, movers, history and a headline summary.
get_top_ratesFreeReturn top SwitchWize rates with freshness and attribution. Savings, CDs, checking and money market answer only from their editor-approved snapshot. A category with no approved release yet returns no ranked rates and says so (approval_status: not_yet_approved) rather than serving unverified observations; approvedOnly makes that requirement explicit. Approved responses include product terms and field evidence; stale or unavailable releases return no ranked rates. detail=full (savings) adds the complete shared build: every listed account marked verified or tracked, best picks, movers, history and a headline summary.
get_transfer_bonuseswrite actionFreeReturn current point-transfer bonuses across tracked bank-to-partner routes, each with a verified timestamp, source URL, and best-ever-for-route when real history exists.
Public scan report
scanner v0.1.10 · 2026-10-15 · same rubric, same numbers if you re-run it
- Code scan2 source files scanned25/25
- Live reliabilityremote reachable in 1977ms20/20
- Tool poisoning16 tool descriptions checked15/15
- Auth qualityopen endpoint exposes 1 write-action tools with no auth3/15
- Maintenancelast push 2 days ago15/15
- Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10
Findings (1)
- highWrite-action tools reachable without authentication
auth.open-write
How SwitchWize MCP compares
Against every graded MCP server, from our own scans and uptime checks. Updated after each scan run.
- Safety score
- 86/100
- higher than 72% of 48,243 servers
- Tools
- 16
- the typical server has 7
- Tools that change things
- 6%
- typical: 0%. Fewer write tools means less an agent can break
- Uptime
- 100%
- of our recent checks; the typical remote server answers 100%
- Response time
- 189 ms
- faster than 68% of 30,027 servers
- GitHub stars
- 1
- more than 65% of 28,706 servers
- Downloads a week
- 54
- more than 43% of 12,306 servers with a package
What its tools let an agent do: 15 read · 1 change
Release history: 3 versions in the registry since 2026-09-09, latest 2026-10-12, typically one every 20 days.
Install SwitchWize MCP in Claude Code, Cursor or VS Code
claude mcp add --transport http switchwize-mcp https://www.switchwize.com/api/mcp
What the publisher says
From the SwitchWize MCP repository's README, as published. We do not edit it. Read it on GitHub
SwitchWize MCP client
Read-only MCP tools for U.S. deposit and mortgage rate observations, savings comparisons, index datasets, and credit-card Reality Scores.
Connect now
For clients supporting Streamable HTTP, use the hosted endpoint:
https://www.switchwize.com/api/mcp
No SwitchWize account or API key is required. The API is rate-limited. Documentation: SwitchWize Developers.
Run the stdio client from source
Use a supported Node.js LTS release and npm:
git clone https://github.com/switchwize/switchwize-mcp.git
cd switchwize-mcp
npm ci --ignore-scripts
npm run build
npm startFor a desktop MCP client, configure command node with the absolute path to dist/index.js as its argument. No environment file or database credentials are needed. SWITCHWIZEAPIURL is an optional API-origin override for developers.
Run from npm
npx -y switchwize-mcp@0.1.0For a desktop MCP client, configure command npx with arguments ["-y", "switchwize-mcp@0.1.0"]. No API key is required.
Available tools
- gettoprates
- gettophysa_rates
- getbankgap
- compareswitchsavings
- getratefreshness
- getindexdataset
- getbankgap_index
- getrealityscore
The client discovers the tool list from /api/public/tools and forwards calls to /api/public/tools/{tool}. It does not initiate financial transactions or request bank credentials. Balance inputs supplied to tools are transmitted to the hosted API; do not include account numbers or other sensitive identifiers.
Data boundaries
Observations may be aggregator-sourced and may be aging. Preserve each response's source dates and freshness labels. Monthly index data is not a current offer. Confirm product terms directly with the provider before making a decision.
Attribute data as SwitchWize (switchwize.com). The MIT license covers this client software, not upstream datasets, third-party materials, or trademarks. See data and service terms.
Package verification
npm run verify:package builds the client and checks the npm tarball against an exact allowlist: LICENSE, README.md, dist/index.js, and package.json. npm pack creates a local tarball; neither command publishes it.
License: MIT. Copyright SW Digital Holdings LLC.
Nothing above is checked by us. What we check is on the safety report.
Grade history
- 2026-10-15restoreB → Ascore 86: Write-action tools reachable without authentication
SwitchWize MCP: common questions
- Is SwitchWize MCP server safe?
- Yes, by our scan: it is graded A (86/100). Read the SwitchWize MCP safety report
- How do I install SwitchWize MCP?
- It runs remotely at www.switchwize.com. Add it to Claude Code, Cursor, VS Code or Claude Desktop with the snippets above, or call it through the mcp.market gateway without installing anything.
- Does SwitchWize MCP need an API key?
- Not as far as the registry entry and our scan can tell: no credentials are declared or required.
- Is SwitchWize MCP maintained?
- The last commit was 24 days ago (2026-10-13). The latest release is v0.3.0.
- Is SwitchWize MCP up?
- 100% of our last 28 checks got an answer. We check remote servers about four times a day.
- How does SwitchWize MCP compare with other MCP servers?
- safety score 86/100 (higher than 72% of 48,243 servers); tools 16 (the typical server has 7); tools that change things 6% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.