@pmelab/gtd 41.0.0 → 43.0.0

package/src/workflows/feature/refactor-review.feature

@@ -78,7 +78,7 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
       """
     And step "refactor.health.check" changes nothing
     And step "refactor.closing" runs its script
-    And step "refactor.build.review.reviewing" writes ".gtd/REVIEW.md" with:
+    And step "refactor.build.review.composing" writes ".gtd/REVIEW.md" with:
       """
       # Review: abc1234
 
@@ -171,7 +171,7 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
       """
     And step "refactor.health.check" changes nothing
     And step "refactor.closing" runs its script
-    And step "refactor.build.review.reviewing" writes ".gtd/REVIEW.md" with:
+    And step "refactor.build.review.composing" writes ".gtd/REVIEW.md" with:
       """
       # Review: abc1234
 
@@ -191,8 +191,8 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
     And the commit subjects end with:
       """
       gtd(check): refactor.closing → refactor.build.quality.correctness.reviewing
-      gtd(agent): refactor.build.quality.correctness.reviewing → refactor.build.review.reviewing
-      gtd(agent): refactor.build.review.reviewing → refactor.build.review.await-review
+      gtd(agent): refactor.build.quality.correctness.reviewing → refactor.build.review.composing
+      gtd(check): refactor.build.review.composing → refactor.build.review.await-review
       """
 
 
@@ -263,7 +263,7 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
         """
       And step "refactor.health.check" changes nothing
       And step "refactor.closing" runs its script
-      And step "refactor.build.review.reviewing" writes ".gtd/REVIEW.md" with:
+      And step "refactor.build.review.composing" writes ".gtd/REVIEW.md" with:
         """
         # Review: abc1234
 
@@ -345,91 +345,10 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
         gtd(check): refactor.deriving → refactor.building
         gtd(agent): refactor.building → refactor.health.check
         gtd(check): refactor.health.check → refactor.closing
-        gtd(check): refactor.closing → refactor.build.review.reviewing
-        gtd(agent): refactor.build.review.reviewing → refactor.build.review.await-review
+        gtd(check): refactor.closing → refactor.build.review.composing
+        gtd(check): refactor.build.review.composing → refactor.build.review.await-review
         """
 
-    Scenario: the refactor review says the whole refactor on its second round
-      Given step "refactor.build.review.reviewing" is prompted with "it covers the whole refactor since it started, on every round"
-      And step "implement.building" writes "src/calc.ts" with:
-        """
-        export const op = (a: number, b: number, sign: number) => a + sign * b
-        export const add = (a: number, b: number) => op(a, b, 1)
-        export const subtract = (a: number, b: number) => op(a, b, -1)
-        """
-      And step "implement.building" writes "src/acceptance.test.ts" with:
-        """
-        it("[AC-1] holds", () => {})
-        """
-      And step "implement.health.check" changes nothing
-      And step "implement.closing" runs its script
-      And judge step "implement.coverage" answers:
-        """
-        [{"id":"AC-1","answer":"yes","p":0.9}]
-        """
-      When the process is driven
-      Then the process rests at "refactor.build.review.await-review"
-      Given a file ".gtd/REVIEW.md" with:
-        """
-        # Review: abc1234
-
-        <!-- base: 0000000 -->
-
-        ## calc
-
-        - [ ] ./src/calc.ts#1-2
-          add built on the shared op[^why]
-
-        [^why]:
-            - H: name the helper `combine`
-        """
-      And step "refactor.build.review.collecting" writes ".gtd/REFACTOR.md" with:
-        """
-        Name the shared helper `combine`.
-        """
-      And step "refactor.build.review.collecting" writes ".gtd/REVIEW.md" with:
-        """
-        # Review: abc1234
-
-        <!-- base: 0000000 -->
-
-        ## calc
-
-        - [ ] ./src/calc.ts#1-2
-          add built on the shared op[^why]
-
-        [^why]:
-            - H: name the helper `combine`
-            - A: noted in the refactor package.
-        """
-      When gtd lands
-      And the process is driven
-      Then the process rests at "refactor.build.review.await-review"
-      Given a file ".gtd/REVIEW.md" with:
-        """
-        # Review: abc1234
-
-        <!-- base: 0000000 -->
-
-        ## calc
-
-        - [ ] ./src/calc.ts#1-2
-          add built on the shared op
-        """
-      And step "refactor.re-unwind" runs its script
-      And step "refactor.deriving" runs its script
-      And step "refactor.building" writes "src/calc.ts" with:
-        """
-        export const combine = (a: number, b: number, sign: number) => a + sign * b
-        export const add = (a: number, b: number) => combine(a, b, 1)
-        """
-      When gtd lands
-      And the process is driven
-      Then the process rests at "refactor.build.review.await-review"
-      When gtd lands
-      And the process is driven
-      Then the process rests at "build.review.reviewing"
-
     Scenario: a question answered at the refactor review reaches the re-derived package as ### Answered Questions
       When the process is driven
       Then the process rests at "refactor.build.review.await-review"
@@ -515,8 +434,8 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
         gtd(check): refactor.deriving → refactor.building
         gtd(agent): refactor.building → refactor.health.check
         gtd(check): refactor.health.check → refactor.closing
-        gtd(check): refactor.closing → refactor.build.review.reviewing
-        gtd(agent): refactor.build.review.reviewing → refactor.build.review.await-review
+        gtd(check): refactor.closing → refactor.build.review.composing
+        gtd(check): refactor.build.review.composing → refactor.build.review.await-review
         """
 
     Scenario: an edit note the collecting turn leaves out refuses the landing and never signs off
@@ -727,7 +646,7 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
         """
       When gtd lands
       And the process is driven
-      Then the process rests at "build.review.reviewing"
+      Then the process rests at "build.review.composing"
       And "src/calc.ts" contains "subtract"
 
     Scenario: a praise-only one-shot footnote signs off and the feature slice follows
@@ -769,7 +688,7 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
         """
       When gtd lands
       And the process is driven
-      Then the process rests at "build.review.reviewing"
+      Then the process rests at "build.review.composing"
       And "src/calc.ts" contains "subtract"
 
     Scenario: a one-shot footnote judged edit that collecting leaves out refuses naming its anchor
@@ -1084,7 +1003,7 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
         """
       When gtd lands
       And the process is driven
-      Then the process rests at "build.review.reviewing"
+      Then the process rests at "build.review.composing"
       And "src/calc.ts" contains "subtract"
 
     Scenario: a praise-only refactor review still signs off and the feature slice follows
@@ -1119,7 +1038,7 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
         """
       When gtd lands
       And the process is driven
-      Then the process rests at "build.review.reviewing"
+      Then the process rests at "build.review.composing"
       And "src/calc.ts" contains "subtract"
 
     Scenario: a nit fix in the refactor slice that changes a non-stray is refused, though a glob stray widens the write access
@@ -1184,8 +1103,8 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
       Then the process rests at "refactor.build.review.await-review"
       And the commit subjects end with:
         """
-        gtd(check): refactor.closing → refactor.build.review.reviewing
-        gtd(agent): refactor.build.review.reviewing → refactor.build.review.await-review
+        gtd(check): refactor.closing → refactor.build.review.composing
+        gtd(check): refactor.build.review.composing → refactor.build.review.await-review
         """
 
   Rule: with a quality lens on both slices
@@ -1256,7 +1175,7 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
         """
       And step "refactor.health.check" changes nothing
       And step "refactor.closing" runs its script
-      And step "refactor.build.review.reviewing" writes ".gtd/REVIEW.md" with:
+      And step "refactor.build.review.composing" writes ".gtd/REVIEW.md" with:
         """
         # Review: abc1234
 
@@ -1309,11 +1228,11 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
       And the resting step's content contains the hash of the commit "gtd(check): refactor.build.review.closing → implement.building"
       Given step "build.quality.correctness.reviewing" changes nothing
       When the process is driven
-      Then the process rests at "build.review.reviewing"
+      Then the process rests at "build.review.composing"
       And the commit subjects end with:
         """
         gtd(check): implement.closing → build.quality.correctness.reviewing
-        gtd(agent): build.quality.correctness.reviewing → build.review.reviewing
+        gtd(agent): build.quality.correctness.reviewing → build.review.composing
         """
 
   Rule: a refactor package never carries a test file, and demoted headings stop at ######
@@ -1463,7 +1382,7 @@ Feature: the refactor slice is reviewed on its own before the feature is impleme
         """
       And step "refactor.health.check" changes nothing
       And step "refactor.closing" runs its script
-      And step "refactor.build.review.reviewing" writes ".gtd/REVIEW.md" with:
+      And step "refactor.build.review.composing" writes ".gtd/REVIEW.md" with:
         """
         # Review: abc1234
 

package/src/workflows/feature/refactor.test.ts

@@ -781,8 +781,6 @@ interface ReviewedScript {
   /** The collecting turn writes no package. */
   readonly silentCollect?: boolean
   readonly collected?: string
-  /** The note the first `review.reviewing` turn writes; later ones write `add`. */
-  readonly firstReviewNote?: string
 }
 
 /** Runs `reviewedRefactor` against a scripted context; a third await-review stops the run. */
@@ -797,7 +795,6 @@ const driveReviewed = async (
   const files = new Map<string, string>()
   let last = ""
   let awaits = 0
-  let reviews = 0
   let reds = script.reds ?? 0
   let checking = false
   let pending: boolean | Error = false
@@ -823,9 +820,7 @@ const driveReviewed = async (
     ...Object.fromEntries(lensEffects),
     deriving: () => void files.set(REFACTOR_PACKAGE, "package"),
     closing: () => void files.delete(REFACTOR_PACKAGE),
-    "review.reviewing": () => {
-      files.set(REVIEW, reviewDoc(reviews++ === 0 ? (script.firstReviewNote ?? "add") : "add"))
-    },
+    "review.composing": () => void files.set(REVIEW, reviewDoc("add")),
     "review.closing": () => void files.delete(REVIEW),
     "review.collecting": () => {
       if (!script.silentCollect)
@@ -846,6 +841,14 @@ const driveReviewed = async (
           if (request.kind === "restart") return Promise.resolve()
           record(request)
           effects[request.name]?.()
+          // Every lens finding is mechanical, so the fix turn runs.
+          if (request.kind === "judge" && request.name === "quality.triage")
+            return Promise.resolve({
+              answers: Object.fromEntries(
+                request.questions.map((q) => [q.id, { answer: "mechanical", p: 0.9 }]),
+              ),
+              truncated: [],
+            })
           return Promise.resolve(
             request.kind === "judge" ? { answers: script.answers ?? {}, truncated: [] } : undefined,
           )
@@ -926,7 +929,7 @@ describe("reviewedRefactor", () => {
       "refactor.build.quality.correctness.reviewing",
       "refactor.build.quality.ponytail-review.reviewing",
       "refactor.build.quality.conventions.reviewing",
-      "refactor.build.review.reviewing",
+      "refactor.build.review.composing",
       "refactor.build.review.await-review",
       "refactor.build.review.closing",
     ])
@@ -934,8 +937,8 @@ describe("reviewedRefactor", () => {
 
   it("reviews from HEAD at scope entry", async () => {
     const h = await driveReviewed(shape())
-    const reviewing = h.steps.find((s) => s.kind === "agent" && s.name === "review.reviewing")
-    expect(reviewing?.kind === "agent" ? reviewing.prompt : "").toContain("base: c0")
+    const composing = h.steps.find((s) => s.kind === "run" && s.name === "review.composing")
+    expect(composing?.kind === "run" ? composing.body : "").toContain("<!-- base: c0 -->")
   })
 
   it("names the entry commit in every lens prompt, on the first lap and after a feedback lap", async () => {
@@ -1192,10 +1195,6 @@ describe("reviewedRefactor's tail prompts", () => {
   it.each([
     ["refactor.build.fix", { findings: true, reds: 2 }],
     ["refactor.build.fix.quality.fixing", { findings: true }],
-    [
-      "refactor.build.review.fix.risks.fixing",
-      { firstReviewNote: "Risk: drops the carry", notes: [undefined] },
-    ],
     [
       "refactor.build.review.fix.nits.fixing",
       { notes: ["nit"], answers: { "note-1": { answer: "nit", p: 0.9 } } },
@@ -1211,17 +1210,3 @@ describe("reviewedRefactor's tail prompts", () => {
     },
   )
 })
-
-describe("reviewedRefactor's review lap wording", () => {
-  it("says the whole refactor since it started on the first and the second round", async () => {
-    const h = await driveReviewed(shape(), { notes: ["add — rename it", undefined] })
-    const prompts = h.steps.flatMap((s) =>
-      s.kind === "agent" && s.name === "review.reviewing" ? [s.prompt] : [],
-    )
-    expect(prompts).toHaveLength(2)
-    for (const prompt of prompts) {
-      expect(prompt).toContain("it covers the whole refactor since it started, on every round")
-      expect(prompt).not.toContain("only what is new since the previous review")
-    }
-  })
-})

package/src/workflows/feature/refactor.ts

@@ -389,7 +389,6 @@ export const reviewedRefactor = async (
         fixAccess: strayAccess(p.strays),
         lensBase: entry,
         rules: t.refactorRules(p.strays, p.pattern),
-        lap: "refactor",
       })
       if (outcome.verdict === "signoff") return true
       await reUnwind(outcome)

package/src/workflows/feature/review-answers.feature

@@ -0,0 +1,179 @@
+@inmem
+Feature: every note in the review is answered with one tick
+
+  Each note gtd writes to `.gtd/REVIEW.md` ends in three boxes — fix as
+  suggested, won't fix, or the human's own answer written over the
+  placeholder. The gate refuses to land until every note holds exactly one
+  tick. A fix goes to the nit batch, a won't-fix is dropped and recorded in
+  the gate's commit, an own answer is judged like any other note. Nothing is
+  deferred to `.gtd/TODO.md`.
+
+  Background:
+    Given the shipped workflows, as gtd install scaffolds them
+    And an environment variable "GTD_QUALITYREVIEWS" set to "correctness"
+    And an environment variable "GTD_FLAGMODE" set to "false"
+    And a file "src/calc.ts" with:
+      """
+      export const add = (a: number, b: number) => a + b
+      """
+    And the working tree is committed as "feat: add calculator"
+    And a file "TODO.md" with:
+      """
+      # Add subtract
+
+      subtract(a, b) returns the difference.
+      """
+    And step "unwind" changes nothing
+    And step "start-gate.check" changes nothing
+    And step "design.triage" writes ".gtd/REQUIREMENTS.md" with:
+      """
+      # Requirements
+
+      ## Concerns
+
+      ### Subtract
+
+      - `subtract(a, b)` returns `a - b`.
+
+      ## Acceptance Criteria
+
+      - AC-1: subtract(a, b) returns a - b.
+      """
+    And step "design.gate.answer" changes nothing
+    And step "seam.naive" writes "src/calc.test.ts" with:
+      """
+      it("[AC-1] subtracts", () => {})
+      """
+    And step "seam.naive" writes "src/calc.ts" with:
+      """
+      export const add = (a: number, b: number) => a + b
+      export const subtract = (a: number, b: number) => a - b
+      """
+    And step "seam.check" changes nothing
+    And step "implement.applying" runs its script
+    And step "implement.health.check" changes nothing
+    And step "implement.closing" runs its script
+    And judge step "implement.coverage" answers:
+      """
+      [{"id":"AC-1","answer":"yes","p":0.9}]
+      """
+    And step "build.quality.correctness.reviewing" writes ".gtd/QUALITY.md" with:
+      """
+      # Quality
+
+      ## subtract has no negative test
+
+      Add a test subtracting a larger number.
+
+      ## add could be a method
+
+      A class would read better.
+
+      ## AC-1 ignores floating point
+
+      Should subtract round? The criterion does not say.
+      """
+    And step "build.quality.sorting" runs its script
+    And step "build.review.composing" runs its script
+    And judge step "build.quality.triage" answers:
+      """
+      [{"id":"Q1","answer":"noise","p":0.9},{"id":"Q2","answer":"noise","p":0.9},{"id":"Q3","answer":"design","p":0.9}]
+      """
+    And step "build.review.closing" runs its script
+
+  Scenario: the note ends in its three answer boxes
+    When the process is driven
+    Then the process rests at "build.review.await-review"
+    And ".gtd/REVIEW.md" contains:
+      """
+      ## AC-1 ignores floating point
+
+      Should subtract round? The criterion does not say.
+
+      Tick exactly one:
+
+      - [ ] fix as suggested
+      - [ ] won't fix
+      - [ ] _your answer_
+      """
+
+  Scenario: a sign-off with the note unanswered is refused
+    When the process is driven
+    Then landing is refused with "answer-completeness: 1 note(s) in .gtd/REVIEW.md not answered"
+    And the process rests at "build.review.await-review"
+
+  Scenario: two ticks on one note are refused
+    When the process is driven
+    And the text "- [ ] fix as suggested" in ".gtd/REVIEW.md" is replaced with "- [x] fix as suggested"
+    And the text "- [ ] won't fix" in ".gtd/REVIEW.md" is replaced with "- [x] won't fix"
+    Then landing is refused with "tick exactly one box per note"
+
+  Scenario: a ticked answer still holding the placeholder is refused
+    When the process is driven
+    And the text "- [ ] _your answer_" in ".gtd/REVIEW.md" is replaced with "- [x] _your answer_"
+    Then landing is refused with "answer-completeness"
+
+  Scenario: won't fix drops the note, records it, and signs off
+    When the process is driven
+    And the text "- [ ] won't fix" in ".gtd/REVIEW.md" is replaced with "- [x] won't fix"
+    And gtd lands
+    And the process is driven
+    Then the process rests at "idle"
+    And the commit "gtd(human): build.review.await-review → build.review.closing" has a body containing "Gtd-Mark: wontfix=AC-1 ignores floating point"
+    And ".gtd/TODO.md" does not exist
+    And no commit body contains "build.review.triage"
+
+  Scenario: fix as suggested goes to the nit batch unjudged, and the note is not asked again
+    Given step "build.review.fix.nits.fixing" writes "src/calc.ts" with:
+      """
+      export const add = (a: number, b: number) => a + b
+      export const subtract = (a: number, b: number) => Math.round(a - b)
+      """
+    And step "build.review.fix.nits.fixing" is prompted with "Should subtract round?"
+    And step "build.health.check" changes nothing
+    When the process is driven
+    And the text "- [ ] fix as suggested" in ".gtd/REVIEW.md" is replaced with "- [x] fix as suggested"
+    And gtd lands
+    And the process is driven
+    Then the process rests at "build.review.await-review"
+    And no commit body contains "build.review.triage"
+    And ".gtd/REVIEW.md" contains "## Fixed from your notes"
+    And ".gtd/REVIEW.md" does not contain "AC-1 ignores floating point"
+    And ".gtd/TODO.md" does not exist
+    When gtd lands
+    And the process is driven
+    Then the process rests at "idle"
+
+  Scenario: an own answer is judged like a note
+    Given judge step "build.review.triage" answers:
+      """
+      [{"id":"answer-1","answer":"nit","p":0.9}]
+      """
+    And step "build.review.fix.nits.fixing" is prompted with "round half up"
+    And step "build.health.check" changes nothing
+    And step "build.review.fix.nits.fixing" writes "src/calc.ts" with:
+      """
+      export const add = (a: number, b: number) => a + b
+      export const subtract = (a: number, b: number) => Math.round(a - b)
+      """
+    When the process is driven
+    And the text "- [ ] _your answer_" in ".gtd/REVIEW.md" is replaced with "- [x] round half up"
+    And gtd lands
+    And the process is driven
+    Then the process rests at "build.review.await-review"
+    And the commit subjects end with:
+      """
+      gtd(human): build.review.await-review → build.review.triage
+      gtd(judge): build.review.triage → build.review.fix.nits.fixing
+      gtd(agent): build.review.fix.nits.fixing → build.health.check
+      gtd(check): build.health.check → build.review.closing
+      gtd(check): build.review.closing → build.review.composing
+      gtd(check): build.review.composing → build.review.await-review
+      """
+    And ".gtd/REVIEW.md" does not contain "AC-1 ignores floating point"
+
+  Scenario: deleting a note's boxes instead of answering is refused
+    When the process is driven
+    And the text "- [ ] fix as suggested" in ".gtd/REVIEW.md" is replaced with ""
+    And the text "Tick exactly one:" in ".gtd/REVIEW.md" is replaced with ""
+    Then landing is refused with "its answer boxes were deleted"

package/src/workflows/feature/review-gate.feature

@@ -57,7 +57,7 @@ Feature: a feature process ends at the human review
       """
       [{"id":"AC-1","answer":"yes","p":0.9}]
       """
-    And step "build.review.reviewing" writes ".gtd/REVIEW.md" with:
+    And step "build.review.composing" writes ".gtd/REVIEW.md" with:
       """
       # Review: abc1234
 
@@ -78,8 +78,8 @@ Feature: a feature process ends at the human review
     And the commit subjects end with:
       """
       gtd(judge): implement.coverage → implement.closing
-      gtd(check): implement.closing → build.review.reviewing
-      gtd(agent): build.review.reviewing → build.review.await-review
+      gtd(check): implement.closing → build.review.composing
+      gtd(check): build.review.composing → build.review.await-review
       """
 
   Scenario: a sign-off ends the process at idle
@@ -102,7 +102,7 @@ Feature: a feature process ends at the human review
     And ".gtd/REVIEW.md" does not exist
     And the commit subjects end with:
       """
-      gtd(agent): build.review.reviewing → build.review.await-review
+      gtd(check): build.review.composing → build.review.await-review
       gtd(human): build.review.await-review → build.review.closing
       gtd(check): build.review.closing → idle
       """
@@ -214,8 +214,8 @@ Feature: a feature process ends at the human review
         gtd(agent): implement.building → implement.health.check
         gtd(check): implement.health.check → implement.coverage
         gtd(judge): implement.coverage → implement.closing
-        gtd(check): implement.closing → build.review.reviewing
-        gtd(agent): build.review.reviewing → build.review.await-review
+        gtd(check): implement.closing → build.review.composing
+        gtd(check): build.review.composing → build.review.await-review
         """
 
     Scenario: a design numbering new behaviour at or below an earlier ID is refused