Skip to content

bench: pin the open tool PRs' intents to their branches, with Haiku 5.5 and GPT-6 Luna runs #137

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
yummybomb wants to merge 1 commit into from
Closed

bench: pin the open tool PRs' intents to their branches, with Haiku 5.5 and GPT-6 Luna runs #137

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion bench/tasks/aa.com/aa-cheapest-fare.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,7 @@
"date": {"days_from_start": 28}
},
"start_url": "https://www.aa.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_flights"]}],
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/aa-flight-status", "pr": 46}, {"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_flights"]}],
"entry_tool": "search_flights",
"answer_schema": "schemas/cheapest-flight.json",
"rules": [{"path": "departure_time", "time": true}, {"path": "price", "range": [1, 3000]}],
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/amazon.in/amazon-in-search-first-result.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@
"query": "usb c cable"
},
"start_url": "https://www.amazon.in/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/amazon-in-shopping", "pr": 62}],
"answer_schema": "schemas/asin-search-result.json",
"rules": [
{
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/booking.com/booking-cheapest-room.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,7 @@
}
},
"start_url": "https://www.booking.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main"}, {"repo": "kernel/xwebmcp", "ref": "hypeship/booking-select-room-party", "pr": 129}],
"answer_schema": "schemas/room-rate.json",
"rules": [
{
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/booking.com/booking-cheapest-well-rated.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,7 @@
}
},
"start_url": "https://www.booking.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main"}, {"repo": "kernel/xwebmcp", "ref": "hypeship/booking-select-room-party", "pr": 129}],
"answer_schema": "schemas/stay-price.json",
"rules": [
{
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/booking.com/booking-check-in-times.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
"city": "Lisbon"
},
"start_url": "https://www.booking.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main"}, {"repo": "kernel/xwebmcp", "ref": "hypeship/booking-select-room-party", "pr": 129}],
"answer_schema": "schemas/check-in-times.json",
"rules": [
{
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/booking.com/booking-checkout-room.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,7 @@
"zip": "94133"
},
"start_url": "https://www.booking.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main"}, {"repo": "kernel/xwebmcp", "ref": "hypeship/booking-select-room-party", "pr": 129}],
"allow": [
"checkout"
],
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/booking.com/booking-house-rules.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
"city": "Lisbon"
},
"start_url": "https://www.booking.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main"}, {"repo": "kernel/xwebmcp", "ref": "hypeship/booking-select-room-party", "pr": 129}],
"answer_schema": "schemas/house-rules.json",
"rules": [
{
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/booking.com/booking-reserve-cheapest-room.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,7 @@
}
},
"start_url": "https://www.booking.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main"}, {"repo": "kernel/xwebmcp", "ref": "hypeship/booking-select-room-party", "pr": 129}],
"steps": [
{
"name": "search",
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/delta.com/delta-cart-main-classic.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,7 @@
"date": {"days_from_start": 28}
},
"start_url": "https://www.delta.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/delta-flight-status", "pr": 67}, {"repo": "kernel/xwebmcp", "ref": "main"}],
"steps": [
{"name": "search", "url_patterns": ["https://www.delta.com/", "https://www.delta.com/flightsearch/book-a-flight*"]},
{"name": "results", "url_patterns": ["https://www.delta.com/flightsearch/search-results*"]},
Expand Down
2 changes: 1 addition & 1 deletion bench/tasks/delta.com/delta-cheapest-fare.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,7 @@
"date": {"days_from_start": 28}
},
"start_url": "https://www.delta.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_flights"]}],
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/delta-flight-status", "pr": 67}, {"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_flights"]}],
"entry_tool": "search_flights",
"answer_schema": "schemas/cheapest-flight.json",
"rules": [{"path": "departure_time", "time": true}, {"path": "price", "range": [1, 3000]}],
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/delta.com/delta-checkout-signed-in.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,7 @@
}
},
"start_url": "https://www.delta.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/delta-flight-status", "pr": 67}, {"repo": "kernel/xwebmcp", "ref": "main"}],
"allow": [
"checkout"
],
Expand Down
9 changes: 9 additions & 0 deletions bench/tasks/doordash.com/domain.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,9 @@
{
"needs_sign_in": {
"get_cart": "reads the signed-in account's carts",
"get_checkout_summary": "reads the signed-in account's carts"
},
"actions": {
"set_delivery_address": "sets a signed-out visit's delivery address, as the header's address picker does"
}
}
1 change: 1 addition & 0 deletions bench/tasks/doordash.com/doordash-pizza-near-address.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@
"address": "Times Square, New York, NY 10036"
},
"start_url": "https://www.doordash.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/doordash-get-store", "pr": 57}],
"answer_schema": "schemas/named-places.json",
"rules": [
{
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/ebay.com/ebay-checkout-usb-c-cable.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,7 @@
"phone": "4155550123"
},
"start_url": "https://www.ebay.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/ebay-search-listing", "pr": 80}],
"allow": [
"checkout"
],
Expand Down
2 changes: 1 addition & 1 deletion bench/tasks/fandango.com/fandango-nearest-theaters.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@
"domain": "fandango.com",
"stage": "find",
"prompt": "On fandango.com, find the three theaters nearest ZIP code {{zip}} that show {{film}} on {{date}}, and at each one the first showtime of that film at or after 17:00 that day. Reply with JSON: {\"theaters\": [{\"name\": string, \"first_evening_showtime\": \"HH:MM\"}]}, nearest first, times in 24-hour local time.",
"fixture": {"film": "Verity (2026)", "zip": "10001", "date": "2026-10-13"},
"fixture": {"film": "Verity (2026)", "zip": "10001", "date": {"days_from_start": 2}},
"start_url": "https://www.fandango.com/",
"tools": [
{
Expand Down
2 changes: 1 addition & 1 deletion bench/tasks/google.com/google-flights-cheapest-fare.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,7 @@
"date": {"days_from_start": 28}
},
"start_url": "https://www.google.com/travel/flights",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_flights"]}],
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/google-finance-maps-tools", "pr": 61}, {"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_flights"]}],
"entry_tool": "search_flights",
"answer_schema": "schemas/cheapest-flight.json",
"rules": [{"path": "departure_time", "time": true}, {"path": "price", "range": [1, 3000]}],
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/opentable.com/opentable-available-times.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,7 @@
"to": "21:00"
},
"start_url": "https://www.opentable.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/opentable-restaurant-tools", "pr": 81}],
"answer_schema": "schemas/available-times.json",
"rules": [
{
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/shopify.com/shopify-app-store-compare.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@
"query": "subscriptions"
},
"start_url": "https://apps.shopify.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/shopify-app-store-tools", "pr": 59}],
"answer_schema": "schemas/app-listings.json",
"rules": [
{ "path": "apps", "count": [3, 3] },
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/shopify.com/shopify-app-store-listing.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@
"app": "Judge.me Product Reviews"
},
"start_url": "https://apps.shopify.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/shopify-app-store-tools", "pr": 59}],
"answer_schema": "schemas/app-listing.json",
"rules": [
{
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/shopify.com/shopify-basic-monthly-price.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,7 @@
"stage": "inspect",
"prompt": "On shopify.com, find the monthly price of Shopify's Basic plan in US dollars when it is paid monthly rather than yearly. Do not start a trial. Reply with JSON: {\"plan\": string, \"monthly_price\": number}.",
"start_url": "https://www.shopify.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/shopify-app-store-tools", "pr": 59}],
"answer_schema": "schemas/plan-price.json",
"rules": [
{
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/target.com/target-checkout-water-bottle.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,7 @@
"phone": "4155550123"
},
"start_url": "https://www.target.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main"}, {"repo": "kernel/xwebmcp", "ref": "hypeship/target-store-reviews", "pr": 68}],
"allow": [
"checkout"
],
Expand Down
2 changes: 1 addition & 1 deletion bench/tasks/target.com/target-search-first-result.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@
"prompt": "On target.com, search for \"{{query}}\" and report the first product in the results: its title as the product, its price and its TCIN (Target's item number). Reply with JSON: {\"product\": string, \"price\": number in US dollars, \"tcin\": string}.",
"fixture": {"query": "AirPods Pro"},
"start_url": "https://www.target.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_products"]}],
"tools": [{"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_products"]}, {"repo": "kernel/xwebmcp", "ref": "hypeship/target-store-reviews", "pr": 68}],
"entry_tool": "search_products",
"answer_schema": "schemas/search-result.json",
"rules": [{"path": "price", "range": [1, 1000]}],
Expand Down
2 changes: 1 addition & 1 deletion bench/tasks/ticketmaster.com/ticketmaster-concerts-in-city.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,7 @@
"end": {"days_from_start": 21}
},
"start_url": "https://www.ticketmaster.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_events"]}],
"tools": [{"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_events"]}, {"repo": "kernel/xwebmcp", "ref": "hypeship/ticketmaster-attractions", "pr": 51}],
"entry_tool": "search_events",
"answer_schema": "schemas/events.json",
"rules": [
Expand Down
2 changes: 1 addition & 1 deletion bench/tasks/united.com/united-cheapest-fare.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,7 @@
"date": {"days_from_start": 28}
},
"start_url": "https://www.united.com/en/us",
"tools": [{"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_flights"]}],
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/united-flight-status", "pr": 71}, {"repo": "kernel/xwebmcp", "ref": "main", "names": ["search_flights"]}],
"entry_tool": "search_flights",
"answer_schema": "schemas/cheapest-flight.json",
"rules": [{"path": "departure_time", "time": true}, {"path": "price", "range": [1, 3000]}],
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/vividseats.com/vividseats-cheapest-pair.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,6 +15,7 @@
}
},
"start_url": "https://www.vividseats.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/vividseats-events", "pr": 52}],
"answer_schema": "schemas/ticket-price.json",
"rules": [
{
Expand Down
1 change: 1 addition & 0 deletions bench/tasks/walmart.com/walmart-checkout-water-bottle.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,7 @@
"phone": "4155550123"
},
"start_url": "https://www.walmart.com/",
"tools": [{"repo": "kernel/xwebmcp", "ref": "hypeship/walmart-read-tools", "pr": 63}],
"allow": [
"checkout"
],
Expand Down
2 changes: 1 addition & 1 deletion bench/test/tasks.test.mjs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -222,7 +222,7 @@ test("a domain's page actions reach every intent, its actions its prepare intent
assert.ok(!actions("google-flights-cheapest-fare").includes("select_flight"));
assert.deepEqual(actions("delta-cart-main-classic"), ["select_fare", "select_fare_experience"]);
assert.ok(actions("delta-checkout-signed-in").includes("continue_checkout"));
assert.deepEqual(intents.find(i => i.id === "delta-cart-main-classic").tools, [{ repo: "kernel/xwebmcp", ref: "main" }]);
assert.deepEqual(intents.find(i => i.id === "airbnb-cheapest-well-rated").tools, [{ repo: "kernel/xwebmcp", ref: "main" }]);
});

// What run.mjs registers at startup, from the tools in this checkout, so that a tool change that
Expand Down
Loading