API & MCP · GA at launch

App intelligence your code and your agent can call.

REST for your scripts, webhooks for your pipelines, MCP for your agent. One contract underneath, so the numbers match everywhere.

DataPussy is API-first. The dashboard is one client among several, and it doesn't get anything the API can't give you. A single contract defines every resource, every filter and every estimate, and the REST API, the OpenAPI spec, the SDK and the MCP server are all generated from it.

REST v1

  • Resources: apps (with estimates, rankings, countries, versions, in-app purchases, reviews and changes), companies, charts, ads, videos, keywords, screens, watchlists, alerts, webhooks, events, exports, meta and usage.
  • App ids look like app_…, and the API also accepts ios:123 and gp:com.example.app.
  • Filters as field[op]=value, or a JSON body on POST /v1/apps/search.
  • fields, expand and view to get exactly the shape you need.
  • Cursor pagination pinned to a snapshot_id, so pages don't shift under you mid-scan.
  • Errors as RFC 9457 problem details, with a type you can switch on.
  • Idempotency-Key on writes, so a retry never does the work twice.
  • dry_run to price a request and max_credits to cap it.
  • X-Credits-* and RateLimit-* headers on every response.
  • p10 and p90 in every view, including the concise one.

Example request and response headers — illustrative, not real data:

curl -X POST "https://api.datapussy.com/v1/apps/search?dry_run=true&max_credits=200" \
  -H "Authorization: Bearer $DATAPUSSY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"filters": {"store": "ios", "country": "US", "category": "health-fitness"}, "view": "concise"}'

# HTTP/1.1 200 OK
# X-Credits-Estimated: 50
# X-Credits-Remaining: 2950

Webhooks and Events

Watchlists and alerts can deliver to email or to your endpoint. Webhook deliveries are signed with HMAC following the Standard Webhooks spec and retried on failure. Every event also lands in GET /v1/events, so a deploy or an outage on your side never means a lost change.

The MCP Server

The MCP server exposes 14 tools, written by hand on top of the same operations as the API:

  • apps_search — filter and sort apps with the same grammar as the API.
  • apps_get — one app: metadata, current estimates and provenance.
  • apps_timeseries — rank, ratings and estimate history over time.
  • charts_get — a store chart for a country, category and date.
  • ads_search — ads by app, country or date (Beta).
  • videos_search — short-form videos linked to apps as mentions (Beta).
  • creators_search — creators derived from our corpus, with attribution confidence (Beta).
  • keywords_analyze — keyword popularity and our own difficulty index.
  • reviews_query — review aggregates and velocity, with review text marked untrusted.
  • screens_search — store screenshots analyzed into structured data (Beta).
  • changes_list — what changed: prices, versions, metadata, ranks and availability.
  • watchlists_manage — create and edit watchlists. The only tool that writes.
  • meta_reference — countries, categories, charts, the filter catalog and method versions.
  • account_usage — your plan, credits left and rate limits.

Each tool returns structuredContent with an outputSchema, plus the cost and the cursor. Results come in a concise view, light on tokens, or a detailed one with full provenance. Both keep P10 and P90. Text written by third parties, such as reviews and ad copy, is marked as untrusted so your agent doesn't take instructions from it. Sign-in uses OAuth 2.1 with PKCE, and the server speaks MCP revision 2026-07-28 with negotiation for older clients.

Plans at Launch

  • Free — $0, 250 credits a month, 10 requests a minute. No card.
  • Indie — $29 a month or $290 a year, 3,000 credits, 60 requests a minute.
  • Team — $99 a month or $990 a year, 15,000 credits, 5 seats, 300 requests a minute.
  • Packs that never expire — 5k credits for $59, 25k for $225, 100k for $700.

API keys start with purr_live_. The CLI is called purr too: purr apps search runs the same query as the API. Refunds within 14 days, and canceling takes two clicks.

For builders

Built for code and agents first

The dashboard, the API and the MCP server run on the same contract and the same filter catalog. Nothing is exclusive to one surface.

REST API v1

Cursor pagination pinned to a snapshot, fields, expand and view to shape responses, RFC 9457 errors, Idempotency-Key, dry_run and max_credits.

MCP server with 14 tools

Curated tools, a concise and a detailed view, cost and cursor in every response, OAuth 2.1 with PKCE, and third-party text marked as untrusted.

Signed webhooks and an events feed

HMAC-signed deliveries following Standard Webhooks, with retries. Missed one? Replay it from GET /v1/events.

OpenAPI 3.1, SDK and CLI

A spec you can lint and generate from, a TypeScript SDK, and purr, a command-line client in Beta.

How it works

From public signal to honest number

Every number goes through the same steps, and every step leaves a trace you can inspect: source, kind, time and method version.

01

Try it without a key

The playground and the demo endpoints answer real requests with no sign-up.

02

Get a free API key

No card required. The Free plan comes with 250 credits a month at launch.

03

Price the call first

Add dry_run to see what a request will cost, and max_credits to cap it. Every response reports what it spent.

04

Connect your agent

Add the MCP server to Claude Code, Cursor or any MCP client. Sign in with OAuth 2.1, or use your API key for headless runs.

05

Subscribe to changes

Point a webhook at your endpoint and let rank, price, version and availability changes come to you.

API & MCP

Built for your code and your agents

A REST API (v1), an MCP server with 14 tools and HMAC-signed webhooks, all on one contract, so the numbers match everywhere. Opens at launch.

See the API & MCP

REST API v1

Filter with the same grammar as the dashboard. Price any call with dry_run, cap it with max_credits.

request.shIllustrative
curl -X POST "https://api.datapussy.com/v1/apps/search?dry_run=true&max_credits=200" \
  -H "Authorization: Bearer $DATAPUSSY_API_KEY" \
  -d '{
    "filters": {
      "store": "ios",
      "country": "US",
      "category": "health-fitness"
    },
    "view": "concise"
  }'

# response headers
HTTP/1.1 200 OK
X-Credits-Estimated: 50

Signed webhooks

Changes delivered to your endpoint, HMAC-signed per Standard Webhooks and retried. Missed one? Replay it from GET /v1/events.

POST → your endpointIllustrative
POST /hooks/datapussy HTTP/1.1
webhook-id: msg_…
webhook-timestamp: 1767225600
webhook-signature: v1,…

{
  "type": "app.rank_changed",
  "app": "ios:…",
  "from": 31,
  "to": 12,
  "country": "US"
}

MCP server, 14 tools

Plug DataPussy into Claude, Cursor or any MCP client. Hand-written tools on the same operations as the API, a concise or a detailed view with P10 and P90 kept in both, and the cost in every answer.

  • apps_search
  • apps_get
  • apps_timeseries
  • changes_list
  • watchlists_manage
  • +9 more
tools/call · apps_getIllustrative
// tool call
{
  "tool": "apps_get",
  "arguments": {
    "id": "ios:…",
    "view": "concise"
  }
}

// result (excerpt)
{
  "downloads_30d": {
    "p10": 340000,
    "p50": 412000,
    "p90": 498000,
    "grade": "B",
    "kind": "modeled",
    "method_version": "v0.3"
  },
  "cost": { "credits": 1 }
}

What every response carries

The same receipts in the API, the MCP server and the webhooks, so a number never travels alone.

  • Estimate bandp10 · p50 · p90
  • Confidence gradegrade: A–D
  • Sourcesource
  • Kindobserved | modeled
  • Observed atobserved_at
  • Method versionmethod_version
  • Credits usedcost.credits · X-Credits-*
  • Cursor + snapshotcursor · snapshot_id

Code is illustrative: hosts, ids and values are placeholders. Access opens at launch.

FAQ

Questions a skeptic would ask

Good. Ask them. Here are the straight answers.

From public signals we collect ourselves: official Apple endpoints and RSS feeds, public store pages, ad libraries and public video pages, read without logging in. Revenue and downloads are then modeled from those signals. Every field says whether we observed it or modeled it, and when.

Still curious?

A person reads every email.

Waitlist

Stop guessing in the dark.

The free plan, the first Data Drops and the private alpha go to the waitlist first.

Product updates only. Unsubscribe anytime. See our privacy notice.