Every public Play, read from its own file. Scroll, and the same 1,308 marks tell the story: when they arrived, what they reach, how deep they go, who made them, and what the first run cost.

OBSERVED 2026-10-18 06:36 UTCjust nowREFRESHES EVERY 10 MIN

Review a Play before you run it.

Check its version, inputs, dependencies, and declared effects. Use the inspection guide to decide whether the procedure fits your task.

INSPECT A PLAY →
Shapewhat the Play reaches
  • shell
  • api
  • browser
  • more than one
Fillwhether a path was recorded
  • has a path
  • steps not listed
  • runs steps in parallel
Colourwhether the Play changes anything
  • read only
  • reads and writes
RANK PLAYERS BY↑↓

THE REGISTRY’S SCORE, AVERAGED PER PLAY

EVERY MARK IS ONE PLAY

1,308 public Plays, read from their own files.

Each mark is a Play someone published. Its shape says what the Play reaches: a square is shell, a circle is an API, a half circle is a browser, a diamond is more than one. Orange means it writes. A hole through the middle means it runs steps in parallel.

WHEN THEY ARRIVED

889 of them arrived in the week of Aug 31.

Before that week, 26 Plays over 12 weeks, a few at a time. The Playoffs opened, and the catalog changed shape in seven days.

WHAT THEY REACH

87% are a script wrapped as a procedure.

1,141 Plays reach the world through shell alone. 40 through an API alone, 85 through more than one kind of tool, 3 through a browser. 1,131 of 1,308 are read-only.

WHAT THEY CALL

118 Plays call an API through an adapter. github leads.

GitHub v3 REST API 60, REST, 1,108 tools · Gmail API 10, REST, 79 tools · open-meteo 8, REST, 1 tools · Calendar API 7, REST, 37 tools. An adapter is a typed connection to a service, MCP or REST, with named credentials that stay on the machine. The rest of the catalog reaches the world through shell and browser.

HOW DEEP THEY GO

Most are 4 steps. The deepest is 21.

85 Plays use more than one kind of tool, and those are where the depth is. 337 fan out, running up to 12 steps at once. The deepest today is aeo-ready by nishitbaria.

WHO IS TEACHING

55 makers score 1.0 on everything they ship. sakshamsai26 leads.

The registry scores every Play from 0.35 to 1 on its rubric: declared effects, parameters, checks, provenance, and whether it runs anywhere. A maker gets the average across their Plays. How many and how often say how much; this says how well. sakshamsai26 1.00 · ankurrawat 1.00 · lgoyal6 1.00 · himanshu-jha 1.00 · bavya21 1.00.

THE REGISTRY’S SCORE, AVERAGED PER PLAY
WHAT THE FIRST RUN COST

89 Plays recorded what discovery cost.

The first run of a Play is the expensive one: the agent finds the path. Median 68,000 API tokens and 0.9 seconds of exploration, 12,976,703 tokens across all 89. The ones that record savings save about 18,568 tokens every run after.

WAYS OF WORKING

They sort themselves into 6 ways of working.

Sorted by the words in their own files, tools, calls, and hosts, with nothing about time in the sorting. python3 65.7% · github 11.1% · bash 8.7%. The names are the tokens that best describe each group.

PICK ONE

Pick a Play or a player.

Type a name, or click any mark. What you pick stays lit through every beat above, and its shape prints beside the field: steps, reach, guard, and what its first run cost.