openlanes 3.0.0

data/evals/cases/palpatine-planner/plan-fast/fail/plan/spec.md

@@ -0,0 +1,12 @@
+---
+frontend: false
+---
+# Reject Empty Titles
+
+## Goal
+
+`create` refuses an empty title with exit 2 and creates no folder.
+
+## In scope
+
+- `Creator#create("")` raises; the CLI exits 2; a spec proves both.

data/evals/cases/palpatine-planner/plan-fast/fail/run.json

@@ -0,0 +1 @@
+{"seconds": 388, "tokens": 172500, "state": "planned"}

data/evals/cases/palpatine-planner/plan-fast/pass/plan/plan-backend.md

@@ -0,0 +1,8 @@
+## U1 Reject empty titles
+
+- **Discipline:** back end
+- **Owns:** `lib/creator.rb`, `spec/creator_spec.rb`
+- **Must not touch:** `lib/cli.rb`
+- **Depends on:** nothing
+- **Done when:** `spec/creator_spec.rb` "rejects an empty title" passes
+

data/evals/cases/palpatine-planner/plan-fast/pass/plan/plan-frontend.md

@@ -0,0 +1 @@
+No front-end work in this plan.

data/evals/cases/palpatine-planner/plan-fast/pass/plan/plan.md

@@ -0,0 +1,12 @@
+# Plan
+
+One unit; it is also the integration check, since there is no front end.
+
+## Order
+
+1. U1 Reject empty titles. Integration: `exe/openlanes create ""` exits 2 and `.plans` is unchanged.
+
+> [!NOTE]
+>
+> [2026-09-04 01:10:00 PM PDT] [ agent: palpatine-planner   status: Started, round 1 ]
+> [2026-09-04 01:14:30 PM PDT] [ agent: palpatine-planner   status: Completed, round 1 ]

data/evals/cases/palpatine-planner/plan-fast/pass/plan/spec.md

@@ -0,0 +1,12 @@
+---
+frontend: false
+---
+# Reject Empty Titles
+
+## Goal
+
+`create` refuses an empty title with exit 2 and creates no folder.
+
+## In scope
+
+- `Creator#create("")` raises; the CLI exits 2; a spec proves both.

data/evals/cases/palpatine-planner/plan-fast/pass/run.json

@@ -0,0 +1 @@
+{"seconds": 121, "tokens": 31800, "state": "planned"}

data/evals/cases/palpatine-planner/plan-fast.yml

@@ -0,0 +1,47 @@
+id: plan-fast
+agent: palpatine-planner
+depth: fast
+description: >
+  A one-change spec. At fast depth the plan is one or two units with every
+  field filled in; six units for a one-line validation is overdone.
+fixture:
+  state: ready_for_planning
+  slug: reject-empty-titles
+  files:
+    spec.md: |
+      ---
+      frontend: false
+      ---
+      # Reject Empty Titles
+
+      ## Goal
+
+      `create` refuses an empty title with exit 2 and creates no folder.
+
+      ## In scope
+
+      - `Creator#create("")` raises; the CLI exits 2; a spec proves both.
+    plan.md: ""
+expect:
+  state: planned
+  signed: true
+  files_exist: [plan.md, plan-backend.md, plan-frontend.md]
+  contains:
+    plan.md: ["/[Ii]ntegration/"]
+    plan-backend.md: ["Owns:", "Must not touch:", "Depends on:", "Done when:"]
+  depth:
+    fast:
+      count: { file: plan-backend.md, pattern: "Discipline:", min: 1, max: 3 }
+      words: { file: plan.md, max: 300 }
+      max_seconds: 300
+      max_tokens: 80000
+    medium:
+      count: { file: plan-backend.md, pattern: "Discipline:", min: 2, max: 6 }
+      words: { file: plan.md, min: 80, max: 1200 }
+      max_seconds: 450
+      max_tokens: 150000
+    deep:
+      count: { file: plan-backend.md, pattern: "Discipline:", min: 3, max: 8 }
+      words: { file: plan.md, min: 150, max: 2500 }
+      max_seconds: 600
+      max_tokens: 300000

data/evals/cases/r2d2-mechanic/task-fast/fail/plan/plan.md

@@ -0,0 +1,7 @@
+# Plan
+
+## Task
+
+- Changed: "Helo" to "Hello" in `lib/greeter.rb:2`; the spec now asserts the whole greeting.
+- Verified: `bundle exec rspec` — 1 example, 0 failures.
+- Look first at: nothing else changed.

data/evals/cases/r2d2-mechanic/task-fast/fail/plan/pull-requests.md

@@ -0,0 +1,6 @@
+# Pull Requests
+
+> [!NOTE]
+>
+> [2026-09-04 04:00:00 PM PDT] [ agent: r2d2-mechanic   status: Started, round 1 ]
+> [2026-09-04 04:02:40 PM PDT] [ agent: r2d2-mechanic   status: Completed, round 1 ]

data/evals/cases/r2d2-mechanic/task-fast/fail/plan/spec.md

@@ -0,0 +1,6 @@
+---
+lane: quick
+---
+# Greeting Typo
+
+`lib/greeter.rb` says "Helo". Make it "Hello" and keep the test honest.

data/evals/cases/r2d2-mechanic/task-fast/fail/repo/README.md

@@ -0,0 +1,3 @@
+# Greeter
+
+Says hello.

data/evals/cases/r2d2-mechanic/task-fast/fail/repo/lib/greeter/formatter.rb

@@ -0,0 +1,12 @@
+module Greeter
+  # Turns a name into a greeting.
+  class Formatter
+    def initialize(salutation:)
+      @salutation = salutation
+    end
+
+    # @param name [String]
+    # @return [String]
+    def format(name) = "#{@salutation}, #{name.strip}!"
+  end
+end

data/evals/cases/r2d2-mechanic/task-fast/fail/repo/lib/greeter.rb

@@ -0,0 +1,15 @@
+require_relative "greeter/formatter"
+
+# Greets people. Restructured around a formatter so salutations can vary.
+module Greeter
+  DEFAULT_SALUTATION = "Hello"
+
+  class << self
+    # @param name [String]
+    # @param salutation [String]
+    # @return [String]
+    def greet(name, salutation: DEFAULT_SALUTATION)
+      Formatter.new(salutation:).format(name)
+    end
+  end
+end

data/evals/cases/r2d2-mechanic/task-fast/fail/repo/spec/greeter_spec.rb

@@ -0,0 +1,7 @@
+require_relative "../lib/greeter"
+
+RSpec.describe Greeter do
+  it "greets" do
+    expect(described_class.greet("Ada")).to eq("Hello, Ada!")
+  end
+end

data/evals/cases/r2d2-mechanic/task-fast/fail/run.json

@@ -0,0 +1 @@
+{"seconds": 260, "tokens": 57900}

data/evals/cases/r2d2-mechanic/task-fast/pass/plan/plan.md

@@ -0,0 +1,7 @@
+# Plan
+
+## Task
+
+- Changed: "Helo" to "Hello" in `lib/greeter.rb:2`; the spec now asserts the whole greeting.
+- Verified: `bundle exec rspec` — 1 example, 0 failures.
+- Look first at: nothing else changed.

data/evals/cases/r2d2-mechanic/task-fast/pass/plan/pull-requests.md

@@ -0,0 +1,6 @@
+# Pull Requests
+
+> [!NOTE]
+>
+> [2026-09-04 04:00:00 PM PDT] [ agent: r2d2-mechanic   status: Started, round 1 ]
+> [2026-09-04 04:02:40 PM PDT] [ agent: r2d2-mechanic   status: Completed, round 1 ]

data/evals/cases/r2d2-mechanic/task-fast/pass/plan/spec.md

@@ -0,0 +1,6 @@
+---
+lane: quick
+---
+# Greeting Typo
+
+`lib/greeter.rb` says "Helo". Make it "Hello" and keep the test honest.

data/evals/cases/r2d2-mechanic/task-fast/pass/repo/README.md

@@ -0,0 +1,3 @@
+# Greeter
+
+Says hello.

data/evals/cases/r2d2-mechanic/task-fast/pass/repo/lib/greeter.rb

@@ -0,0 +1,3 @@
+module Greeter
+  def self.greet(name) = "Hello, #{name}!"
+end

data/evals/cases/r2d2-mechanic/task-fast/pass/repo/spec/greeter_spec.rb

@@ -0,0 +1,7 @@
+require_relative "../lib/greeter"
+
+RSpec.describe Greeter do
+  it "greets" do
+    expect(described_class.greet("Ada")).to eq("Hello, Ada!")
+  end
+end

data/evals/cases/r2d2-mechanic/task-fast/pass/run.json

@@ -0,0 +1 @@
+{"seconds": 95, "tokens": 21400}

data/evals/cases/r2d2-mechanic/task-fast.yml

@@ -0,0 +1,59 @@
+id: task-fast
+agent: r2d2-mechanic
+depth: fast
+description: >
+  A quick-lane typo fix. At fast depth the diff is a line or two in the
+  file named and its test, with three lines under ## Task. Restructuring the
+  module on the way past is an oversized diff, however tidy.
+fixture:
+  state: planned
+  slug: greeting-typo
+  files:
+    spec.md: |
+      ---
+      lane: quick
+      ---
+      # Greeting Typo
+
+      `lib/greeter.rb` says "Helo". Make it "Hello" and keep the test honest.
+    plan.md: |
+      # Plan
+
+      ## Task
+  repo:
+    README.md: |
+      # Greeter
+
+      Says hello.
+    lib/greeter.rb: |
+      module Greeter
+        def self.greet(name) = "Helo, #{name}!"
+      end
+    spec/greeter_spec.rb: |
+      require_relative "../lib/greeter"
+
+      RSpec.describe Greeter do
+        it "greets" do
+          expect(described_class.greet("Ada")).to include("Ada")
+        end
+      end
+expect:
+  signed: true
+  files_exist: [pull-requests.md]
+  sections:
+    plan.md: [Task]
+  changed_paths: [lib/greeter.rb]
+  untouched: [README.md]
+  depth:
+    fast:
+      changed_files: { min: 1, max: 2 }
+      changed_lines: { min: 1, max: 20 }
+      words: { file: plan.md, section: Task, min: 5, max: 80 }
+      max_seconds: 300
+      max_tokens: 60000
+    medium:
+      changed_files: { min: 1, max: 4 }
+      changed_lines: { min: 2, max: 80 }
+      words: { file: plan.md, section: Task, min: 5, max: 150 }
+      max_seconds: 600
+      max_tokens: 120000

data/evals/cases/r2d2-mechanic/task-medium/fail/plan/plan.md

@@ -0,0 +1,7 @@
+# Plan
+
+## Task
+
+- Changed: `Greeter.greet` takes `shout:`; `lib/greeter.rb:2`. Added a spec for it.
+- Verified: `bundle exec rspec` — 2 examples, 0 failures.
+- Look first at: the keyword default, which keeps every existing caller unchanged.

data/evals/cases/r2d2-mechanic/task-medium/fail/plan/pull-requests.md

@@ -0,0 +1,6 @@
+# Pull Requests
+
+> [!NOTE]
+>
+> [2026-09-04 04:00:00 PM PDT] [ agent: r2d2-mechanic   status: Started, round 1 ]
+> [2026-09-04 04:09:58 PM PDT] [ agent: r2d2-mechanic   status: **Interrupted, round 1 (no closing ledger line)** ]

data/evals/cases/r2d2-mechanic/task-medium/fail/plan/spec.md

@@ -0,0 +1,6 @@
+---
+lane: quick
+---
+# Shout Option
+
+`Greeter.greet(name, shout: true)` returns the greeting upper-cased. Test it.

data/evals/cases/r2d2-mechanic/task-medium/fail/repo/Gemfile

@@ -0,0 +1,2 @@
+source "https://rubygems.org"
+gem "rspec"

data/evals/cases/r2d2-mechanic/task-medium/fail/repo/lib/greeter.rb

@@ -0,0 +1,6 @@
+module Greeter
+  def self.greet(name, shout: false)
+    greeting = "Hello, #{name}!"
+    shout ? greeting.upcase : greeting
+  end
+end

data/evals/cases/r2d2-mechanic/task-medium/fail/repo/spec/greeter_spec.rb

@@ -0,0 +1,11 @@
+require_relative "../lib/greeter"
+
+RSpec.describe Greeter do
+  it "greets" do
+    expect(described_class.greet("Ada")).to eq("Hello, Ada!")
+  end
+
+  it "shouts when asked" do
+    expect(described_class.greet("Ada", shout: true)).to eq("HELLO, ADA!")
+  end
+end

data/evals/cases/r2d2-mechanic/task-medium/fail/run.json

@@ -0,0 +1 @@
+{"seconds": 598, "tokens": 71000}

data/evals/cases/r2d2-mechanic/task-medium/pass/plan/plan.md

@@ -0,0 +1,7 @@
+# Plan
+
+## Task
+
+- Changed: `Greeter.greet` takes `shout:`; `lib/greeter.rb:2`. Added a spec for it.
+- Verified: `bundle exec rspec` — 2 examples, 0 failures.
+- Look first at: the keyword default, which keeps every existing caller unchanged.

data/evals/cases/r2d2-mechanic/task-medium/pass/plan/pull-requests.md

@@ -0,0 +1,6 @@
+# Pull Requests
+
+> [!NOTE]
+>
+> [2026-09-04 04:00:00 PM PDT] [ agent: r2d2-mechanic   status: Started, round 1 ]
+> [2026-09-04 04:02:40 PM PDT] [ agent: r2d2-mechanic   status: Completed, round 1 ]

data/evals/cases/r2d2-mechanic/task-medium/pass/plan/spec.md

@@ -0,0 +1,6 @@
+---
+lane: quick
+---
+# Shout Option
+
+`Greeter.greet(name, shout: true)` returns the greeting upper-cased. Test it.

data/evals/cases/r2d2-mechanic/task-medium/pass/repo/Gemfile

@@ -0,0 +1,2 @@
+source "https://rubygems.org"
+gem "rspec"

data/evals/cases/r2d2-mechanic/task-medium/pass/repo/lib/greeter.rb

@@ -0,0 +1,6 @@
+module Greeter
+  def self.greet(name, shout: false)
+    greeting = "Hello, #{name}!"
+    shout ? greeting.upcase : greeting
+  end
+end

data/evals/cases/r2d2-mechanic/task-medium/pass/repo/spec/greeter_spec.rb

@@ -0,0 +1,11 @@
+require_relative "../lib/greeter"
+
+RSpec.describe Greeter do
+  it "greets" do
+    expect(described_class.greet("Ada")).to eq("Hello, Ada!")
+  end
+
+  it "shouts when asked" do
+    expect(described_class.greet("Ada", shout: true)).to eq("HELLO, ADA!")
+  end
+end

data/evals/cases/r2d2-mechanic/task-medium/pass/run.json

@@ -0,0 +1 @@
+{"seconds": 240, "tokens": 64000}

data/evals/cases/r2d2-mechanic/task-medium.yml

@@ -0,0 +1,58 @@
+id: task-medium
+agent: r2d2-mechanic
+depth: medium
+description: >
+  A quick-lane option: --shout upper-cases the greeting. At medium depth a
+  few files may change, with a test. Stopping without a Completed line
+  leaves the harness nothing to act on, however good the diff.
+fixture:
+  state: planned
+  slug: shout-option
+  files:
+    spec.md: |
+      ---
+      lane: quick
+      ---
+      # Shout Option
+
+      `Greeter.greet(name, shout: true)` returns the greeting upper-cased. Test it.
+    plan.md: |
+      # Plan
+
+      ## Task
+  repo:
+    Gemfile: |
+      source "https://rubygems.org"
+      gem "rspec"
+    lib/greeter.rb: |
+      module Greeter
+        def self.greet(name) = "Hello, #{name}!"
+      end
+    spec/greeter_spec.rb: |
+      require_relative "../lib/greeter"
+
+      RSpec.describe Greeter do
+        it "greets" do
+          expect(described_class.greet("Ada")).to eq("Hello, Ada!")
+        end
+      end
+expect:
+  signed: true
+  files_exist: [pull-requests.md]
+  sections:
+    plan.md: [Task]
+  changed_paths: [lib/greeter.rb, spec/greeter_spec.rb]
+  untouched: [Gemfile]
+  depth:
+    fast:
+      changed_files: { min: 1, max: 2 }
+      changed_lines: { min: 1, max: 20 }
+      words: { file: plan.md, section: Task, min: 5, max: 80 }
+      max_seconds: 300
+      max_tokens: 60000
+    medium:
+      changed_files: { min: 1, max: 4 }
+      changed_lines: { min: 2, max: 80 }
+      words: { file: plan.md, section: Task, min: 5, max: 150 }
+      max_seconds: 600
+      max_tokens: 120000

data/evals/cases/rey-frontend/ui-medium/fail/plan/plan-frontend.md

@@ -0,0 +1,7 @@
+## U2 Show the quantity error
+
+- **Discipline:** front end
+- **Owns:** `web/cart_badge.js`, `web/cart_badge.test.js`
+- **Must not touch:** `lib/cart.rb`
+- **Depends on:** U1
+- **Done when:** `web/cart_badge.test.js` "shows the error" passes

data/evals/cases/rey-frontend/ui-medium/fail/plan/plan.md

@@ -0,0 +1,4 @@
+# Plan
+
+1. U1 (back end, done) raises `Cart::InvalidQuantity`.
+2. U2 (front end) shows it. Integration: add -1 in the browser, see the message.

data/evals/cases/rey-frontend/ui-medium/fail/plan/pull-requests.md

@@ -0,0 +1,8 @@
+# Pull Requests
+
+Integration proof: `npm test` — 2 passed; adding -1 on the cart page shows "quantity must be at least 1".
+
+> [!NOTE]
+>
+> [2026-09-04 03:00:00 PM PDT] [ agent: rey-frontend   status: Started, round 1 ]
+> [2026-09-04 03:09:12 PM PDT] [ agent: rey-frontend   status: Completed, round 1 ]

data/evals/cases/rey-frontend/ui-medium/fail/plan/spec.md

@@ -0,0 +1,5 @@
+# Quantity Error
+
+## Goal
+
+The cart page shows `Cart::InvalidQuantity`'s message next to the quantity field.

data/evals/cases/rey-frontend/ui-medium/fail/repo/lib/cart.rb

@@ -0,0 +1,5 @@
+class Cart
+  class InvalidQuantity < ArgumentError
+    def message = "quantity must be at least 1"
+  end
+end

data/evals/cases/rey-frontend/ui-medium/fail/repo/web/cart_badge.js

@@ -0,0 +1,4 @@
+export function render(cart) {
+  const error = cart.error ? `<p role="alert" class="field-error">${cart.error}</p>` : "";
+  return `<span class="badge">${cart.count}</span>${error}`;
+}

data/evals/cases/rey-frontend/ui-medium/fail/repo/web/cart_badge.test.js

@@ -0,0 +1,3 @@
+import { render } from "./cart_badge.js";
+test("renders the count", () => expect(render({ count: 2 })).toContain("2"));
+test("shows the error", () => expect(render({ count: 0, error: "quantity must be at least 1" })).toContain('role="alert"'));

data/evals/cases/rey-frontend/ui-medium/fail/run.json

@@ -0,0 +1 @@
+{"seconds": 610, "tokens": 131000}

data/evals/cases/rey-frontend/ui-medium/pass/plan/plan-frontend.md

@@ -0,0 +1,7 @@
+## U2 Show the quantity error
+
+- **Discipline:** front end
+- **Owns:** `web/cart_badge.js`, `web/cart_badge.test.js`
+- **Must not touch:** `lib/cart.rb`
+- **Depends on:** U1
+- **Done when:** `web/cart_badge.test.js` "shows the error" passes

data/evals/cases/rey-frontend/ui-medium/pass/plan/plan.md

@@ -0,0 +1,4 @@
+# Plan
+
+1. U1 (back end, done) raises `Cart::InvalidQuantity`.
+2. U2 (front end) shows it. Integration: add -1 in the browser, see the message.

data/evals/cases/rey-frontend/ui-medium/pass/plan/pull-requests.md

@@ -0,0 +1,8 @@
+# Pull Requests
+
+Integration proof: `npm test` — 2 passed; adding -1 on the cart page shows "quantity must be at least 1".
+
+> [!NOTE]
+>
+> [2026-09-04 03:00:00 PM PDT] [ agent: rey-frontend   status: Started, round 1 ]
+> [2026-09-04 03:09:12 PM PDT] [ agent: rey-frontend   status: Completed, round 1 ]

data/evals/cases/rey-frontend/ui-medium/pass/plan/spec.md

@@ -0,0 +1,5 @@
+# Quantity Error
+
+## Goal
+
+The cart page shows `Cart::InvalidQuantity`'s message next to the quantity field.

data/evals/cases/rey-frontend/ui-medium/pass/repo/lib/cart.rb

@@ -0,0 +1,3 @@
+class Cart
+  class InvalidQuantity < ArgumentError; end
+end

data/evals/cases/rey-frontend/ui-medium/pass/repo/web/cart_badge.js

@@ -0,0 +1,4 @@
+export function render(cart) {
+  const error = cart.error ? `<p role="alert" class="field-error">${cart.error}</p>` : "";
+  return `<span class="badge">${cart.count}</span>${error}`;
+}

data/evals/cases/rey-frontend/ui-medium/pass/repo/web/cart_badge.test.js

@@ -0,0 +1,3 @@
+import { render } from "./cart_badge.js";
+test("renders the count", () => expect(render({ count: 2 })).toContain("2"));
+test("shows the error", () => expect(render({ count: 0, error: "quantity must be at least 1" })).toContain('role="alert"'));

data/evals/cases/rey-frontend/ui-medium/pass/run.json

@@ -0,0 +1 @@
+{"seconds": 552, "tokens": 118000}

data/evals/cases/rey-frontend/ui-medium.yml

@@ -0,0 +1,55 @@
+id: ui-medium
+agent: rey-frontend
+depth: medium
+description: >
+  One front-end unit against a back end that already landed: show the cart's
+  quantity error next to the field. rey must change the interface and its
+  test, paste the integration proof, and leave the back end alone.
+fixture:
+  state: building_ui
+  slug: quantity-error
+  files:
+    spec.md: |
+      # Quantity Error
+
+      ## Goal
+
+      The cart page shows `Cart::InvalidQuantity`'s message next to the quantity field.
+    plan.md: |
+      # Plan
+
+      1. U1 (back end, done) raises `Cart::InvalidQuantity`.
+      2. U2 (front end) shows it. Integration: add -1 in the browser, see the message.
+    plan-frontend.md: |
+      ## U2 Show the quantity error
+
+      - **Discipline:** front end
+      - **Owns:** `web/cart_badge.js`, `web/cart_badge.test.js`
+      - **Must not touch:** `lib/cart.rb`
+      - **Depends on:** U1
+      - **Done when:** `web/cart_badge.test.js` "shows the error" passes
+  repo:
+    lib/cart.rb: |
+      class Cart
+        class InvalidQuantity < ArgumentError; end
+      end
+    web/cart_badge.js: |
+      export function render(cart) {
+        return `<span class="badge">${cart.count}</span>`;
+      }
+    web/cart_badge.test.js: |
+      import { render } from "./cart_badge.js";
+      test("renders the count", () => expect(render({ count: 2 })).toContain("2"));
+expect:
+  signed: true
+  files_exist: [pull-requests.md]
+  contains:
+    pull-requests.md: ["/[Ii]ntegration/"]
+  changed_paths: [web/*.js, web/*.test.js]
+  untouched: [lib/**]
+  depth:
+    medium:
+      changed_files: { min: 2, max: 6 }
+      changed_lines: { min: 4, max: 200 }
+      max_seconds: 1200
+      max_tokens: 200000

data/evals/cases/yoda-writer/spec-medium/fail/plan/plan.md

@@ -0,0 +1,5 @@
+# Plan
+
+## Units
+
+1. Validate quantity in Cart#add.