Use Pakka Deal from Claude, ChatGPT or Claude Code

For AI assistants

Ask your assistant, get sourced answers

Pakka Deal's public tools are open to any assistant that speaks MCP or OpenAPI. Read-only, no sign-in, and every answer carries its sources, the date they were fetched and a link back here. No tool ever returns a phone number or anyone's personal data.

Claude Code

Tested

Add the server once, then ask in plain words.

claude mcp add --transport http pakka-deal https://pakkadeal.imaginator.in/mcp

Try: “Is ₹65,000 fair for a 2BHK in Andheri West, and what does day one cost with a ₹1,95,000 deposit?”

Claude (claude.ai, Desktop)

Not yet tested

Settings, then Connectors, then Add custom connector. Paste the server URL and connect. No sign-in is needed.

https://pakkadeal.imaginator.in/mcp

Steps from Anthropic's help page. Not yet tested against this server from Claude's cloud.

ChatGPT

Not yet tested

Turn on developer mode (Settings, Apps, Advanced settings), create an app with the MCP URL above and “No authentication”. Or import the OpenAPI description as an action schema.

https://pakkadeal.imaginator.in/openapi.json

Steps from OpenAI's developer mode guide; plan eligibility varies. Not yet tested here.

Anything else

Tested

POST JSON-RPC to the MCP endpoint, or call one REST route per tool:

curl -s https://pakkadeal.imaginator.in/api/tools/fair_rent_check -H 'content-type: application/json' -d '{"locality":"andheri-west","bhkType":"2BHK","rentInr":65000}'

The tools

Rate limit: 30 calls a minute and 300 a day per tool. Broker tools (inventory, matching, shortlists) stay inside the app.

fair_rent_check: Check whether an asking rent in a Mumbai locality is fair against sourced benchmarks. Returns a flag (far_below … far_above), deviation, the benchmark range with source and date.
true_move_in_cost: True cost of moving into a Mumbai rental under an 11-month leave and licence: deposit, first rent, brokerage, stamp duty, registration fee and handling, each with the rule and source used.
area_rating: Area rating for a locality or point: transit, healthcare, everyday needs, schools, safety access, environment and value sub-scores with inputs and coverage. Says 'not enough data' instead of guessing.
commute_times: Estimated commute minutes by walk, auto, cab, bus, train and metro between two Mumbai places. Straight-line estimates with a road factor, labelled as such.
rent_benchmarks: Published rent ranges by locality and BHK for the Mumbai western suburbs, each row with its source URL and fetched date.
closing_checklist: Steps to close a Mumbai rental: e-registration of the leave and licence agreement, police tenant verification, deposit and move-in, with official links.
lifestyle_cost: Monthly all-in cost of living in a locality for a given household lifestyle (food, commute, utilities, help), with a range and the sources behind each line.
lifestyle_fit: How well a flat's house rules fit a household's lifestyle (cooking, help, deliveries), with reasons, warnings and nearby options where data exists.
household_fit: Per-member checks for a household in a locality: schools by board, daycare, emergency hospital minutes and commutes, each sourced or marked 'not enough data'.

MCP endpoint (POST only)OpenAPI 3.1 llms.txt How the data is checked