repull 0.2.7 → 0.2.9

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (285) hide show
  1. checksums.yaml +4 -4
  2. data/lib/repull/api/airbnb_api.rb +117 -130
  3. data/lib/repull/api/atlas_api.rb +1 -1
  4. data/lib/repull/api/billing_api.rb +1 -56
  5. data/lib/repull/api/booking_com_api.rb +20 -119
  6. data/lib/repull/api/connect_api.rb +1 -1
  7. data/lib/repull/api/conversations_api.rb +1 -1
  8. data/lib/repull/api/guests_api.rb +1 -1
  9. data/lib/repull/api/kv_api.rb +1 -1
  10. data/lib/repull/api/listings_api.rb +212 -1
  11. data/lib/repull/api/markets_api.rb +1 -1
  12. data/lib/repull/api/plumguide_api.rb +1 -1
  13. data/lib/repull/api/pricing_api.rb +1 -1
  14. data/lib/repull/api/properties_api.rb +1 -1
  15. data/lib/repull/api/reservations_api.rb +1 -198
  16. data/lib/repull/api/reviews_api.rb +1 -1
  17. data/lib/repull/api/{availability_api.rb → sandbox_api.rb} +36 -67
  18. data/lib/repull/api/schema_api.rb +1 -1
  19. data/lib/repull/api/system_api.rb +1 -1
  20. data/lib/repull/api/vrbo_api.rb +1 -127
  21. data/lib/repull/api/webhooks_api.rb +1 -1
  22. data/lib/repull/api_client.rb +1 -1
  23. data/lib/repull/api_error.rb +1 -1
  24. data/lib/repull/api_model_base.rb +1 -1
  25. data/lib/repull/configuration.rb +1 -1
  26. data/lib/repull/models/account_created_event.rb +1 -1
  27. data/lib/repull/models/account_created_payload.rb +1 -1
  28. data/lib/repull/models/account_disconnected_event.rb +1 -1
  29. data/lib/repull/models/account_disconnected_payload.rb +1 -1
  30. data/lib/repull/models/ai_operation.rb +1 -1
  31. data/lib/repull/models/ai_operation_completed_event.rb +1 -1
  32. data/lib/repull/models/ai_operation_completed_payload.rb +1 -1
  33. data/lib/repull/models/ai_operation_failed_event.rb +1 -1
  34. data/lib/repull/models/ai_operation_failed_payload.rb +1 -1
  35. data/lib/repull/models/ai_operation_failed_payload_error.rb +1 -1
  36. data/lib/repull/models/{create_studio_project_request.rb → airbnb_availability_write_request.rb} +39 -53
  37. data/lib/repull/models/{create_studio_deployment201_response_data.rb → airbnb_calendar_operation.rb} +109 -58
  38. data/lib/repull/models/airbnb_connection.rb +1 -1
  39. data/lib/repull/models/airbnb_connection_accessibility_amenities_inner.rb +1 -1
  40. data/lib/repull/models/airbnb_connection_amenities_inner.rb +1 -1
  41. data/lib/repull/models/airbnb_connection_host.rb +1 -1
  42. data/lib/repull/models/airbnb_connection_response.rb +1 -1
  43. data/lib/repull/models/airbnb_connection_summary.rb +1 -1
  44. data/lib/repull/models/airbnb_data_freshness.rb +1 -1
  45. data/lib/repull/models/airbnb_listing.rb +1 -1
  46. data/lib/repull/models/{update_studio_project_request.rb → airbnb_listing_action_request.rb} +56 -48
  47. data/lib/repull/models/airbnb_listing_list_response.rb +1 -1
  48. data/lib/repull/models/{update_availability_request.rb → airbnb_pricing_write_request.rb} +101 -12
  49. data/lib/repull/models/airbnb_reservation.rb +1 -1
  50. data/lib/repull/models/airbnb_reservation_list_response.rb +1 -1
  51. data/lib/repull/models/airbnb_review.rb +1 -1
  52. data/lib/repull/models/airbnb_review_list_response.rb +1 -1
  53. data/lib/repull/models/airbnb_thread.rb +1 -1
  54. data/lib/repull/models/airbnb_thread_list_response.rb +1 -1
  55. data/lib/repull/models/{create_studio_project201_response_data.rb → booking_availability_update.rb} +126 -49
  56. data/lib/repull/models/{list_studio_deployments200_response.rb → booking_availability_update_request.rb} +91 -19
  57. data/lib/repull/models/{generate_studio_completion_request_project_id.rb → booking_availability_update_request_property_id.rb} +3 -3
  58. data/lib/repull/models/{get_studio_deployment200_response.rb → booking_availability_update_request_updates_inner.rb} +75 -118
  59. data/lib/repull/models/booking_connect_listing_option.rb +1 -1
  60. data/lib/repull/models/booking_connect_room.rb +1 -1
  61. data/lib/repull/models/booking_connect_rooms_response.rb +1 -1
  62. data/lib/repull/models/booking_conversation.rb +1 -1
  63. data/lib/repull/models/booking_conversation_list_response.rb +1 -1
  64. data/lib/repull/models/booking_pricing_rate_update.rb +2 -1
  65. data/lib/repull/models/booking_pricing_rate_update_date_range.rb +1 -1
  66. data/lib/repull/models/booking_pricing_rate_update_restrictions.rb +66 -7
  67. data/lib/repull/models/booking_pricing_response.rb +1 -1
  68. data/lib/repull/models/booking_pricing_update_request.rb +1 -1
  69. data/lib/repull/models/booking_pricing_update_response.rb +1 -1
  70. data/lib/repull/models/booking_property.rb +1 -1
  71. data/lib/repull/models/booking_property_list_response.rb +1 -1
  72. data/lib/repull/models/booking_room_mapping.rb +1 -1
  73. data/lib/repull/models/booking_verify_hotel_request.rb +1 -1
  74. data/lib/repull/models/booking_verify_hotel_response.rb +1 -1
  75. data/lib/repull/models/bulk_pricing_failure.rb +1 -1
  76. data/lib/repull/models/bulk_pricing_item.rb +1 -1
  77. data/lib/repull/models/bulk_pricing_request.rb +1 -1
  78. data/lib/repull/models/bulk_pricing_response.rb +1 -1
  79. data/lib/repull/models/calendar_day.rb +1 -1
  80. data/lib/repull/models/calendar_response.rb +1 -1
  81. data/lib/repull/models/calendar_updated_event.rb +1 -1
  82. data/lib/repull/models/calendar_updated_payload.rb +1 -1
  83. data/lib/repull/models/calendar_updated_payload_range.rb +1 -1
  84. data/lib/repull/models/clear_kv200_response.rb +1 -1
  85. data/lib/repull/models/connect_host.rb +1 -1
  86. data/lib/repull/models/connect_provider.rb +1 -1
  87. data/lib/repull/models/connect_provider_list_response.rb +1 -1
  88. data/lib/repull/models/connect_session.rb +1 -1
  89. data/lib/repull/models/connect_status.rb +1 -1
  90. data/lib/repull/models/connection.rb +1 -1
  91. data/lib/repull/models/connection_list_response.rb +1 -1
  92. data/lib/repull/models/conversation.rb +1 -1
  93. data/lib/repull/models/conversation_detail.rb +1 -1
  94. data/lib/repull/models/conversation_guest.rb +1 -1
  95. data/lib/repull/models/conversation_guest_contact.rb +1 -1
  96. data/lib/repull/models/conversation_host.rb +1 -1
  97. data/lib/repull/models/conversation_list_response.rb +1 -1
  98. data/lib/repull/models/conversation_message_attachment.rb +1 -1
  99. data/lib/repull/models/create_billing_checkout_request.rb +1 -1
  100. data/lib/repull/models/create_connect_session_request.rb +1 -1
  101. data/lib/repull/models/create_connection_request.rb +1 -1
  102. data/lib/repull/models/create_webhook_request.rb +1 -1
  103. data/lib/repull/models/custom_schema.rb +1 -1
  104. data/lib/repull/models/custom_schema_create.rb +1 -1
  105. data/lib/repull/models/custom_schema_create_response.rb +1 -1
  106. data/lib/repull/models/custom_schema_delete_response.rb +1 -1
  107. data/lib/repull/models/custom_schema_list_response.rb +1 -1
  108. data/lib/repull/models/custom_schema_summary.rb +1 -1
  109. data/lib/repull/models/custom_schema_update.rb +1 -1
  110. data/lib/repull/models/delete_kv200_response.rb +1 -1
  111. data/lib/repull/models/error.rb +1 -1
  112. data/lib/repull/models/error_error.rb +1 -1
  113. data/lib/repull/models/error_error_support.rb +1 -1
  114. data/lib/repull/models/get_health200_response.rb +1 -1
  115. data/lib/repull/models/guest.rb +1 -1
  116. data/lib/repull/models/guest_contact.rb +1 -1
  117. data/lib/repull/models/guest_flag.rb +1 -1
  118. data/lib/repull/models/guest_list_response.rb +1 -1
  119. data/lib/repull/models/guest_note.rb +1 -1
  120. data/lib/repull/models/guest_profile.rb +1 -1
  121. data/lib/repull/models/guest_reservations_summary.rb +1 -1
  122. data/lib/repull/models/list_kv200_response.rb +1 -1
  123. data/lib/repull/models/list_kv200_response_data_inner.rb +1 -1
  124. data/lib/repull/models/list_kv200_response_pagination.rb +1 -1
  125. data/lib/repull/models/listing.rb +1 -1
  126. data/lib/repull/models/{create_studio_project_generation_request.rb → listing_active_request.rb} +21 -21
  127. data/lib/repull/models/{delete_studio_project200_response_data.rb → listing_active_response.rb} +13 -11
  128. data/lib/repull/models/listing_address.rb +1 -1
  129. data/lib/repull/models/listing_amenity.rb +1 -1
  130. data/lib/repull/models/listing_channel.rb +1 -1
  131. data/lib/repull/models/listing_comp.rb +1 -1
  132. data/lib/repull/models/listing_comp_nightly.rb +1 -1
  133. data/lib/repull/models/listing_comp_ratings.rb +1 -1
  134. data/lib/repull/models/listing_comps_response.rb +1 -1
  135. data/lib/repull/models/listing_content.rb +1 -1
  136. data/lib/repull/models/{get_studio_project200_response.rb → listing_content_update_request.rb} +107 -11
  137. data/lib/repull/models/{upsert_studio_project_file200_response.rb → listing_content_update_request_address.rb} +54 -11
  138. data/lib/repull/models/listing_content_update_request_amenities.rb +105 -0
  139. data/lib/repull/models/{create_studio_deployment_request.rb → listing_content_update_request_amenities_one_of_inner.rb} +52 -21
  140. data/lib/repull/models/{update_reservation_request.rb → listing_content_update_request_occupancy.rb} +34 -29
  141. data/lib/repull/{api/ai_api.rb → models/listing_content_update_request_photos_inner.rb} +78 -60
  142. data/lib/repull/models/{generate_studio_completion200_response_data.rb → listing_content_update_request_photos_inner_one_of.rb} +82 -67
  143. data/lib/repull/models/{generate_studio_completion200_response.rb → listing_content_update_request_policies.rb} +117 -11
  144. data/lib/repull/models/{create_ai_operation200_response.rb → listing_content_update_response.rb} +33 -17
  145. data/lib/repull/models/listing_create_request.rb +1 -1
  146. data/lib/repull/models/listing_create_response.rb +1 -1
  147. data/lib/repull/models/listing_created_event.rb +1 -1
  148. data/lib/repull/models/listing_created_payload.rb +1 -1
  149. data/lib/repull/models/listing_created_payload_address.rb +1 -1
  150. data/lib/repull/models/listing_deleted_event.rb +1 -1
  151. data/lib/repull/models/listing_deleted_payload.rb +1 -1
  152. data/lib/repull/models/listing_details.rb +1 -1
  153. data/lib/repull/models/listing_generate_content_request.rb +1 -1
  154. data/lib/repull/models/listing_generate_content_response.rb +1 -1
  155. data/lib/repull/models/listing_list_response.rb +1 -1
  156. data/lib/repull/models/listing_pricing_apply_request.rb +1 -1
  157. data/lib/repull/models/listing_pricing_apply_response.rb +1 -1
  158. data/lib/repull/models/listing_pricing_history_entry.rb +1 -1
  159. data/lib/repull/models/listing_pricing_history_response.rb +1 -1
  160. data/lib/repull/models/listing_pricing_recommendation.rb +1 -1
  161. data/lib/repull/models/listing_pricing_response.rb +1 -1
  162. data/lib/repull/models/listing_pricing_response_comp_summary.rb +1 -1
  163. data/lib/repull/models/listing_pricing_response_date_range.rb +1 -1
  164. data/lib/repull/models/listing_pricing_response_listing.rb +1 -1
  165. data/lib/repull/models/listing_pricing_strategy.rb +1 -1
  166. data/lib/repull/models/listing_pricing_strategy_input.rb +1 -1
  167. data/lib/repull/models/listing_publish_airbnb_request.rb +1 -1
  168. data/lib/repull/models/listing_publish_response.rb +1 -1
  169. data/lib/repull/models/listing_publish_status_channel.rb +1 -1
  170. data/lib/repull/models/listing_publish_status_connection.rb +1 -1
  171. data/lib/repull/models/listing_publish_status_response.rb +1 -1
  172. data/lib/repull/models/listing_quality_tier.rb +1 -1
  173. data/lib/repull/models/listing_segment.rb +1 -1
  174. data/lib/repull/models/listing_segment_recommendation.rb +1 -1
  175. data/lib/repull/models/listing_segments_response.rb +1 -1
  176. data/lib/repull/models/listing_segments_response_scope.rb +1 -1
  177. data/lib/repull/models/listing_updated_event.rb +1 -1
  178. data/lib/repull/models/listing_updated_payload.rb +1 -1
  179. data/lib/repull/models/map_airbnb_listing_request.rb +215 -0
  180. data/lib/repull/models/{create_reservation_request.rb → map_airbnb_listing_response.rb} +138 -117
  181. data/lib/repull/models/map_connect_booking_rooms_request.rb +1 -1
  182. data/lib/repull/models/map_connect_booking_rooms_response.rb +1 -1
  183. data/lib/repull/models/market_browse_category.rb +1 -1
  184. data/lib/repull/models/market_browse_entry.rb +1 -1
  185. data/lib/repull/models/market_browse_featured.rb +1 -1
  186. data/lib/repull/models/market_browse_response.rb +1 -1
  187. data/lib/repull/models/market_calendar_day.rb +1 -1
  188. data/lib/repull/models/market_calendar_day_events_inner.rb +1 -1
  189. data/lib/repull/models/market_calendar_response.rb +1 -1
  190. data/lib/repull/models/market_detail_response.rb +1 -1
  191. data/lib/repull/models/market_detail_response_price_distribution_inner.rb +1 -1
  192. data/lib/repull/models/market_detail_response_property_type_mix_inner.rb +1 -1
  193. data/lib/repull/models/market_detail_response_supply_trend_inner.rb +1 -1
  194. data/lib/repull/models/market_detail_response_top_comps.rb +1 -1
  195. data/lib/repull/models/market_event.rb +1 -1
  196. data/lib/repull/models/market_my_listing.rb +1 -1
  197. data/lib/repull/models/market_summary.rb +1 -1
  198. data/lib/repull/models/market_top_comp.rb +1 -1
  199. data/lib/repull/models/markets_overview_response.rb +1 -1
  200. data/lib/repull/models/markets_overview_response_browse.rb +1 -1
  201. data/lib/repull/models/markets_overview_response_subscriptions.rb +1 -1
  202. data/lib/repull/models/markets_overview_response_totals.rb +1 -1
  203. data/lib/repull/models/message.rb +1 -1
  204. data/lib/repull/models/message_list_response.rb +1 -1
  205. data/lib/repull/models/pagination.rb +1 -1
  206. data/lib/repull/models/payment_completed_event.rb +1 -1
  207. data/lib/repull/models/payment_completed_payload.rb +1 -1
  208. data/lib/repull/models/payment_refunded_event.rb +1 -1
  209. data/lib/repull/models/payment_refunded_payload.rb +1 -1
  210. data/lib/repull/models/plumguide_listing.rb +1 -1
  211. data/lib/repull/models/plumguide_listing_list_response.rb +1 -1
  212. data/lib/repull/models/property.rb +43 -70
  213. data/lib/repull/models/property_list_response.rb +1 -1
  214. data/lib/repull/models/reply_booking_review200_response.rb +1 -1
  215. data/lib/repull/models/reply_booking_review_request.rb +1 -1
  216. data/lib/repull/models/repull_ping_event.rb +1 -1
  217. data/lib/repull/models/repull_ping_payload.rb +1 -1
  218. data/lib/repull/models/reservation.rb +29 -2
  219. data/lib/repull/models/reservation_cancelled_event.rb +1 -1
  220. data/lib/repull/models/reservation_cancelled_payload.rb +1 -1
  221. data/lib/repull/models/reservation_created_event.rb +1 -1
  222. data/lib/repull/models/reservation_created_payload.rb +1 -1
  223. data/lib/repull/models/reservation_financials.rb +1 -1
  224. data/lib/repull/models/reservation_list_response.rb +1 -1
  225. data/lib/repull/models/reservation_message_received_event.rb +1 -1
  226. data/lib/repull/models/reservation_message_received_payload.rb +1 -1
  227. data/lib/repull/models/reservation_message_received_payload_from.rb +1 -1
  228. data/lib/repull/models/reservation_occupancy.rb +1 -1
  229. data/lib/repull/models/reservation_primary_guest.rb +1 -1
  230. data/lib/repull/models/reservation_updated_event.rb +1 -1
  231. data/lib/repull/models/reservation_updated_payload.rb +1 -1
  232. data/lib/repull/models/reservation_webhook_object.rb +1 -1
  233. data/lib/repull/models/respond_airbnb_review_request.rb +1 -1
  234. data/lib/repull/models/review.rb +1 -1
  235. data/lib/repull/models/review_category.rb +1 -1
  236. data/lib/repull/models/review_list_response.rb +1 -1
  237. data/lib/repull/models/review_response.rb +1 -1
  238. data/lib/repull/models/rotate_webhook_secret200_response.rb +1 -1
  239. data/lib/repull/models/{upsert_studio_project_file_request.rb → sandbox_fixture_ref.rb} +49 -21
  240. data/lib/repull/models/{delete_studio_deployment200_response_data.rb → sandbox_reset_result.rb} +73 -12
  241. data/lib/repull/models/sandbox_reset_result_deleted.rb +216 -0
  242. data/lib/repull/models/sandbox_seed_result.rb +278 -0
  243. data/lib/repull/models/select_connect_provider_request.rb +1 -1
  244. data/lib/repull/models/select_provider_response.rb +1 -1
  245. data/lib/repull/models/set_kv_request.rb +1 -1
  246. data/lib/repull/models/studio_deployment.rb +1 -1
  247. data/lib/repull/models/studio_error.rb +1 -1
  248. data/lib/repull/models/studio_error_error.rb +1 -1
  249. data/lib/repull/models/studio_file.rb +1 -1
  250. data/lib/repull/models/studio_generation.rb +1 -1
  251. data/lib/repull/models/studio_project.rb +1 -1
  252. data/lib/repull/models/test_webhook_request.rb +1 -1
  253. data/lib/repull/models/update_listing_pricing_strategy200_response.rb +1 -1
  254. data/lib/repull/models/update_webhook_request.rb +1 -1
  255. data/lib/repull/models/vrbo_listing.rb +1 -1
  256. data/lib/repull/models/vrbo_listing_list_response.rb +1 -1
  257. data/lib/repull/models/vrbo_reservation.rb +1 -1
  258. data/lib/repull/models/vrbo_reservation_list_response.rb +1 -1
  259. data/lib/repull/models/webhook_delivery.rb +1 -1
  260. data/lib/repull/models/webhook_delivery_detail.rb +1 -1
  261. data/lib/repull/models/webhook_delivery_list_response.rb +1 -1
  262. data/lib/repull/models/webhook_event.rb +1 -1
  263. data/lib/repull/models/webhook_event_catalog.rb +1 -1
  264. data/lib/repull/models/webhook_event_catalog_domains_inner.rb +1 -1
  265. data/lib/repull/models/webhook_event_catalog_entry.rb +1 -1
  266. data/lib/repull/models/webhook_event_type.rb +1 -1
  267. data/lib/repull/models/webhook_list_response.rb +1 -1
  268. data/lib/repull/models/webhook_subscription.rb +1 -1
  269. data/lib/repull/version.rb +2 -2
  270. data/lib/repull.rb +27 -36
  271. data/openapi/v1.json +3604 -4346
  272. metadata +28 -37
  273. data/lib/repull/api/studio_api.rb +0 -1094
  274. data/lib/repull/models/create_studio_deployment201_response.rb +0 -147
  275. data/lib/repull/models/create_studio_project201_response.rb +0 -147
  276. data/lib/repull/models/create_studio_project_generation201_response.rb +0 -147
  277. data/lib/repull/models/create_studio_project_generation201_response_data.rb +0 -165
  278. data/lib/repull/models/delete_studio_deployment200_response.rb +0 -147
  279. data/lib/repull/models/delete_studio_project200_response.rb +0 -147
  280. data/lib/repull/models/delete_studio_project_file200_response.rb +0 -147
  281. data/lib/repull/models/delete_studio_project_file200_response_data.rb +0 -156
  282. data/lib/repull/models/generate_studio_completion_request.rb +0 -305
  283. data/lib/repull/models/list_studio_project_files200_response.rb +0 -149
  284. data/lib/repull/models/list_studio_projects200_response.rb +0 -149
  285. data/lib/repull/models/upsert_studio_project_file200_response_data.rb +0 -147
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai
@@ -14,14 +14,14 @@ require 'date'
14
14
  require 'time'
15
15
 
16
16
  module Repull
17
- class CreateStudioProjectGenerationRequest < ApiModelBase
18
- # Prompt to send to Repull AI.
19
- attr_accessor :prompt
17
+ class ListingActiveRequest < ApiModelBase
18
+ # Target active state. `false` deactivates (excludes) the listing; `true` reactivates it (subject to the plan-listings cap).
19
+ attr_accessor :active
20
20
 
21
21
  # Attribute mapping from ruby-style variable name to JSON key.
22
22
  def self.attribute_map
23
23
  {
24
- :'prompt' => :'prompt'
24
+ :'active' => :'active'
25
25
  }
26
26
  end
27
27
 
@@ -38,7 +38,7 @@ module Repull
38
38
  # Attribute type mapping.
39
39
  def self.openapi_types
40
40
  {
41
- :'prompt' => :'String'
41
+ :'active' => :'Boolean'
42
42
  }
43
43
  end
44
44
 
@@ -52,22 +52,22 @@ module Repull
52
52
  # @param [Hash] attributes Model attributes in the form of hash
53
53
  def initialize(attributes = {})
54
54
  if (!attributes.is_a?(Hash))
55
- fail ArgumentError, "The input argument (attributes) must be a hash in `Repull::CreateStudioProjectGenerationRequest` initialize method"
55
+ fail ArgumentError, "The input argument (attributes) must be a hash in `Repull::ListingActiveRequest` initialize method"
56
56
  end
57
57
 
58
58
  # check to see if the attribute exists and convert string to symbol for hash key
59
59
  acceptable_attribute_map = self.class.acceptable_attribute_map
60
60
  attributes = attributes.each_with_object({}) { |(k, v), h|
61
61
  if (!acceptable_attribute_map.key?(k.to_sym))
62
- fail ArgumentError, "`#{k}` is not a valid attribute in `Repull::CreateStudioProjectGenerationRequest`. Please check the name to make sure it's valid. List of attributes: " + acceptable_attribute_map.keys.inspect
62
+ fail ArgumentError, "`#{k}` is not a valid attribute in `Repull::ListingActiveRequest`. Please check the name to make sure it's valid. List of attributes: " + acceptable_attribute_map.keys.inspect
63
63
  end
64
64
  h[k.to_sym] = v
65
65
  }
66
66
 
67
- if attributes.key?(:'prompt')
68
- self.prompt = attributes[:'prompt']
67
+ if attributes.key?(:'active')
68
+ self.active = attributes[:'active']
69
69
  else
70
- self.prompt = nil
70
+ self.active = nil
71
71
  end
72
72
  end
73
73
 
@@ -76,8 +76,8 @@ module Repull
76
76
  def list_invalid_properties
77
77
  warn '[DEPRECATED] the `list_invalid_properties` method is obsolete'
78
78
  invalid_properties = Array.new
79
- if @prompt.nil?
80
- invalid_properties.push('invalid value for "prompt", prompt cannot be nil.')
79
+ if @active.nil?
80
+ invalid_properties.push('invalid value for "active", active cannot be nil.')
81
81
  end
82
82
 
83
83
  invalid_properties
@@ -87,18 +87,18 @@ module Repull
87
87
  # @return true if the model is valid
88
88
  def valid?
89
89
  warn '[DEPRECATED] the `valid?` method is obsolete'
90
- return false if @prompt.nil?
90
+ return false if @active.nil?
91
91
  true
92
92
  end
93
93
 
94
94
  # Custom attribute writer method with validation
95
- # @param [Object] prompt Value to be assigned
96
- def prompt=(prompt)
97
- if prompt.nil?
98
- fail ArgumentError, 'prompt cannot be nil'
95
+ # @param [Object] active Value to be assigned
96
+ def active=(active)
97
+ if active.nil?
98
+ fail ArgumentError, 'active cannot be nil'
99
99
  end
100
100
 
101
- @prompt = prompt
101
+ @active = active
102
102
  end
103
103
 
104
104
  # Checks equality by comparing each attribute.
@@ -106,7 +106,7 @@ module Repull
106
106
  def ==(o)
107
107
  return true if self.equal?(o)
108
108
  self.class == o.class &&
109
- prompt == o.prompt
109
+ active == o.active
110
110
  end
111
111
 
112
112
  # @see the `==` method
@@ -118,7 +118,7 @@ module Repull
118
118
  # Calculates hash code according to all attributes.
119
119
  # @return [Integer] Hash code
120
120
  def hash
121
- [prompt].hash
121
+ [active].hash
122
122
  end
123
123
 
124
124
  # Builds the object from hash
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai
@@ -14,16 +14,18 @@ require 'date'
14
14
  require 'time'
15
15
 
16
16
  module Repull
17
- class DeleteStudioProject200ResponseData < ApiModelBase
17
+ class ListingActiveResponse < ApiModelBase
18
+ # The listing id (id fields are serialized as strings to preserve precision).
18
19
  attr_accessor :id
19
20
 
20
- attr_accessor :deleted
21
+ # The resulting active state after the toggle.
22
+ attr_accessor :active
21
23
 
22
24
  # Attribute mapping from ruby-style variable name to JSON key.
23
25
  def self.attribute_map
24
26
  {
25
27
  :'id' => :'id',
26
- :'deleted' => :'deleted'
28
+ :'active' => :'active'
27
29
  }
28
30
  end
29
31
 
@@ -41,7 +43,7 @@ module Repull
41
43
  def self.openapi_types
42
44
  {
43
45
  :'id' => :'String',
44
- :'deleted' => :'Boolean'
46
+ :'active' => :'Boolean'
45
47
  }
46
48
  end
47
49
 
@@ -55,14 +57,14 @@ module Repull
55
57
  # @param [Hash] attributes Model attributes in the form of hash
56
58
  def initialize(attributes = {})
57
59
  if (!attributes.is_a?(Hash))
58
- fail ArgumentError, "The input argument (attributes) must be a hash in `Repull::DeleteStudioProject200ResponseData` initialize method"
60
+ fail ArgumentError, "The input argument (attributes) must be a hash in `Repull::ListingActiveResponse` initialize method"
59
61
  end
60
62
 
61
63
  # check to see if the attribute exists and convert string to symbol for hash key
62
64
  acceptable_attribute_map = self.class.acceptable_attribute_map
63
65
  attributes = attributes.each_with_object({}) { |(k, v), h|
64
66
  if (!acceptable_attribute_map.key?(k.to_sym))
65
- fail ArgumentError, "`#{k}` is not a valid attribute in `Repull::DeleteStudioProject200ResponseData`. Please check the name to make sure it's valid. List of attributes: " + acceptable_attribute_map.keys.inspect
67
+ fail ArgumentError, "`#{k}` is not a valid attribute in `Repull::ListingActiveResponse`. Please check the name to make sure it's valid. List of attributes: " + acceptable_attribute_map.keys.inspect
66
68
  end
67
69
  h[k.to_sym] = v
68
70
  }
@@ -71,8 +73,8 @@ module Repull
71
73
  self.id = attributes[:'id']
72
74
  end
73
75
 
74
- if attributes.key?(:'deleted')
75
- self.deleted = attributes[:'deleted']
76
+ if attributes.key?(:'active')
77
+ self.active = attributes[:'active']
76
78
  end
77
79
  end
78
80
 
@@ -97,7 +99,7 @@ module Repull
97
99
  return true if self.equal?(o)
98
100
  self.class == o.class &&
99
101
  id == o.id &&
100
- deleted == o.deleted
102
+ active == o.active
101
103
  end
102
104
 
103
105
  # @see the `==` method
@@ -109,7 +111,7 @@ module Repull
109
111
  # Calculates hash code according to all attributes.
110
112
  # @return [Integer] Hash code
111
113
  def hash
112
- [id, deleted].hash
114
+ [id, active].hash
113
115
  end
114
116
 
115
117
  # Builds the object from hash
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai
@@ -1,7 +1,7 @@
1
1
  =begin
2
2
  #Repull API
3
3
 
4
- #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 5 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
4
+ #The unified API for vacation rental tech. Connect to 50+ PMS platforms and 4 OTA channels through one REST API. Built-in AI operations for guest communication, pricing, and listing optimization. ## Designed for AI agents Every error response on this API includes machine-parseable fields so an LLM (Claude in MCP, Cursor, Cline, GPT, etc.) can self-recover without escalating to a human: - `error.code` — stable string identifier (e.g. `invalid_params`, `rate_limit_exceeded`) - `error.message` — human-readable cause - `error.fix` — exact recovery steps (e.g. \"Pass `check_in_after` as ISO 8601: `?check_in_after=2026-01-15`\") - `error.docs_url` — link to the canonical write-up at `https://repull.dev/docs/errors/{code}` - `error.request_id` — id to correlate with server-side logs - `error.field` / `error.value_received` / `error.valid_values` / `error.did_you_mean` — when the error is parameter-specific - `error.retry_after` — seconds to wait before retrying (rate-limit + transient upstream) `Access-Control-Expose-Headers` lists `x-request-id` and the `X-RateLimit-*` family so browsers can read them on cross-origin responses. ## Quick Start 1. Get an API key at https://repull.dev/dashboard 2. Connect a PMS: `POST /v1/connect/{provider}` 3. List properties: `GET /v1/properties` 4. Get reservations: `GET /v1/reservations` ## Authentication All requests require a Bearer token: ``` Authorization: Bearer sk_test_YOUR_API_KEY ``` Sandbox keys start with `sk_test_`, production with `sk_live_`. ## Request Correlation (X-Request-ID) Every response carries an `X-Request-ID` header, e.g. `X-Request-ID: req_01HXY...`. Include this id in support tickets and bug reports — we can trace the full request lifecycle (auth, rate limit, handler, downstream calls, log row) from a single id. You may set the header on the inbound request to forward your own trace id; we will echo it back instead of generating a new one. Accepted format: `^[\\\\w.-]{1,128}$`. The id is also embedded in error envelopes as `request_id` so server-side log diffs work even when the response headers are stripped by an intermediate proxy. ## Rate Limits The public API enforces a per-API-key sliding-window rate limit on top of the per-tier monthly + daily-AI quotas. **Default policy:** 600 requests per 60 seconds, per API key. Sliding window — there is no fixed-minute boundary you can burst across. Every response includes: | Header | Meaning | |---|---| | `X-RateLimit-Limit` | Requests permitted in the current window. | | `X-RateLimit-Remaining` | Requests left in the current window after this call. | | `X-RateLimit-Reset` | Unix epoch (seconds) when the next slot opens. | | `X-RateLimit-Policy` | Machine-readable policy descriptor, e.g. `600;w=60`. | | `Retry-After` | Seconds to wait before retrying. **Only present on 429 responses.** | **On 429 (rate_limit_exceeded):** the response body matches the standard error envelope with `code: \"rate_limit_exceeded\"`, plus `limit`, `window_seconds`, `retry_after`, and `request_id` fields. SDKs MUST honor `Retry-After` and use exponential backoff with jitter on subsequent retries — never a tight loop. Recommended backoff: ``` sleep_ms = (Retry-After * 1000) + random(0..250) ``` Monthly + daily-AI tier quotas (`free`, `starter`, `custom`) are enforced separately and also surface as 429s; they include `tier`, `scope`, and `resets_at` fields. ## Plan Limits (402 — `listings_limit_exceeded`) The Repull API also enforces a per-tier cap on **active listings**: | Tier | Active listings cap | |---|---| | `free` | 3 | | `starter` | 50 | | `custom` | unlimited | When a customer's active-listing count is above their tier cap, the API returns **`402 Payment Required`** with `error.code = \"listings_limit_exceeded\"` on every route EXCEPT: - `/v1/health` — uptime probes are never gated. - `/v1/usage/*` — so dashboards can render the over-cap state. - Any `DELETE` — so the customer can trim listings to get back under the cap without paying. Unlike 429, 402 is NOT a \"wait and retry\" condition — `Retry-After` is not set. The only paths back to 200 are: 1. `DELETE` enough listings to come back under the cap, or 2. Upgrade at `https://repull.dev/dashboard/billing`. The server-side usage cache is 60s, so the first 200 after an upgrade may take up to a minute. The envelope mirrors `rate_limit_exceeded` for SDK ergonomics: `tier`, `limit`, `active_listings`, `upgrade_url`, plus the standard `code` / `message` / `fix` / `docs_url` / `request_id`.
5
5
 
6
6
  The version of the OpenAPI document: 1.0.0
7
7
  Contact: ivan@vanio.ai