gsd-pi 2.33.1 → 2.34.0-dev.bbb5216

package/dist/resources/extensions/gsd/tests/progress-score.test.ts

@@ -1,206 +0,0 @@
-/**
- * progress-score.test.ts — Tests for progress score / traffic light (#1221).
- *
- * Tests:
- *   - Score computation from health signals
- *   - Signal evaluation (trend, error streak, recent errors)
- *   - Context-aware scoring (retry counts, unit progress)
- *   - Formatting (single-line, detailed report)
- */
-
-import {
-  recordHealthSnapshot,
-  resetProactiveHealing,
-} from "../doctor-proactive.ts";
-
-import {
-  computeProgressScore,
-  computeProgressScoreWithContext,
-  formatProgressLine,
-  formatProgressReport,
-} from "../progress-score.ts";
-
-import { createTestContext } from "./test-helpers.ts";
-
-const { assertEq, assertTrue, assertMatch, report } = createTestContext();
-
-async function main(): Promise<void> {
-  try {
-    // ── Base Score: No Data ─────────────────────────────────────────────
-    console.log("\n=== progress: green with no data ===");
-    {
-      resetProactiveHealing();
-      const score = computeProgressScore();
-      assertEq(score.level, "green", "green when no data available");
-      assertTrue(score.summary.includes("Progressing well"), "summary says progressing");
-      assertTrue(score.signals.length > 0, "has signals");
-    }
-
-    // ── Green: Clean Health Data ────────────────────────────────────────
-    console.log("\n=== progress: green with clean health ===");
-    {
-      resetProactiveHealing();
-      for (let i = 0; i < 5; i++) {
-        recordHealthSnapshot(0, 0, 0);
-      }
-      const score = computeProgressScore();
-      assertEq(score.level, "green", "green with all clean snapshots");
-    }
-
-    // ── Yellow: Some Warnings ──────────────────────────────────────────
-    console.log("\n=== progress: yellow with error streak ===");
-    {
-      resetProactiveHealing();
-      recordHealthSnapshot(1, 2, 0);
-      recordHealthSnapshot(1, 1, 0);
-      const score = computeProgressScore();
-      assertEq(score.level, "yellow", "yellow with consecutive errors");
-      assertTrue(score.summary.includes("Struggling"), "summary says struggling");
-    }
-
-    // ── Red: Degrading Health ──────────────────────────────────────────
-    console.log("\n=== progress: red with degrading trend ===");
-    {
-      resetProactiveHealing();
-      // 5 older clean snapshots
-      for (let i = 0; i < 5; i++) {
-        recordHealthSnapshot(0, 0, 0);
-      }
-      // 5 recent error snapshots — triggers degrading trend
-      for (let i = 0; i < 5; i++) {
-        recordHealthSnapshot(3, 5, 0);
-      }
-      const score = computeProgressScore();
-      assertEq(score.level, "red", "red with degrading trend and persistent errors");
-      assertTrue(score.summary.includes("Stuck"), "summary says stuck");
-    }
-
-    // ── Red: High Error Streak ─────────────────────────────────────────
-    console.log("\n=== progress: red with high error streak ===");
-    {
-      resetProactiveHealing();
-      for (let i = 0; i < 4; i++) {
-        recordHealthSnapshot(2, 0, 0);
-      }
-      const score = computeProgressScore();
-      assertEq(score.level, "red", "red with 4 consecutive error units");
-    }
-
-    // ── Context-Aware Scoring ──────────────────────────────────────────
-    console.log("\n=== progress: context with retries ===");
-    {
-      resetProactiveHealing();
-      for (let i = 0; i < 3; i++) {
-        recordHealthSnapshot(0, 0, 0);
-      }
-      const score = computeProgressScoreWithContext({
-        currentUnitId: "M001/S01/T03",
-        completedUnits: 2,
-        totalUnits: 5,
-        retryCount: 0,
-        maxRetries: 5,
-      });
-      assertEq(score.level, "green", "green with no retries");
-      assertTrue(score.summary.includes("M001/S01/T03"), "summary includes unit ID");
-      assertTrue(score.summary.includes("2 of 5"), "summary includes progress");
-    }
-
-    console.log("\n=== progress: context with high retry count ===");
-    {
-      resetProactiveHealing();
-      for (let i = 0; i < 3; i++) {
-        recordHealthSnapshot(0, 0, 0);
-      }
-      const score = computeProgressScoreWithContext({
-        currentUnitId: "M001/S01/T03",
-        retryCount: 4,
-        maxRetries: 5,
-      });
-      assertEq(score.level, "red", "red with high retry count");
-      assertTrue(score.summary.includes("looping"), "summary mentions looping");
-    }
-
-    console.log("\n=== progress: context with moderate retries ===");
-    {
-      resetProactiveHealing();
-      for (let i = 0; i < 3; i++) {
-        recordHealthSnapshot(0, 0, 0);
-      }
-      const score = computeProgressScoreWithContext({
-        currentUnitId: "M001/S01/T03",
-        retryCount: 1,
-        maxRetries: 5,
-      });
-      assertEq(score.level, "yellow", "yellow with 1 retry");
-    }
-
-    // ── Formatting ─────────────────────────────────────────────────────
-    console.log("\n=== progress: formatProgressLine ===");
-    {
-      resetProactiveHealing();
-      const score = computeProgressScore();
-      const line = formatProgressLine(score);
-      assertTrue(line.includes("Progressing well"), "line includes summary");
-      // Should start with green circle emoji
-      assertTrue(line.startsWith("\uD83D\uDFE2"), "starts with green circle");
-    }
-
-    console.log("\n=== progress: formatProgressLine yellow ===");
-    {
-      resetProactiveHealing();
-      recordHealthSnapshot(1, 0, 0);
-      recordHealthSnapshot(1, 0, 0);
-      const score = computeProgressScore();
-      const line = formatProgressLine(score);
-      assertTrue(line.startsWith("\uD83D\uDFE1"), "starts with yellow circle");
-    }
-
-    console.log("\n=== progress: formatProgressReport ===");
-    {
-      resetProactiveHealing();
-      recordHealthSnapshot(0, 1, 0);
-      const score = computeProgressScore();
-      const detailed = formatProgressReport(score);
-      assertTrue(detailed.includes("Signals:"), "report has signals section");
-      assertTrue(detailed.includes("health_trend"), "report includes trend signal");
-      assertTrue(detailed.includes("error_streak"), "report includes streak signal");
-    }
-
-    // ── Signal Details ─────────────────────────────────────────────────
-    console.log("\n=== progress: signal names are consistent ===");
-    {
-      resetProactiveHealing();
-      recordHealthSnapshot(0, 0, 0);
-      const score = computeProgressScore();
-      const names = score.signals.map(s => s.name);
-      assertTrue(names.includes("health_trend"), "has health_trend signal");
-      assertTrue(names.includes("error_streak"), "has error_streak signal");
-      assertTrue(names.includes("recent_errors"), "has recent_errors signal");
-      assertTrue(names.includes("artifact_production"), "has artifact_production signal");
-      assertTrue(names.includes("dispatch_velocity"), "has dispatch_velocity signal");
-    }
-
-    console.log("\n=== progress: all signals have valid levels ===");
-    {
-      resetProactiveHealing();
-      for (let i = 0; i < 5; i++) {
-        recordHealthSnapshot(1, 1, 1);
-      }
-      const score = computeProgressScore();
-      for (const signal of score.signals) {
-        assertTrue(
-          signal.level === "green" || signal.level === "yellow" || signal.level === "red",
-          `signal ${signal.name} has valid level: ${signal.level}`,
-        );
-        assertTrue(signal.detail.length > 0, `signal ${signal.name} has non-empty detail`);
-      }
-    }
-
-  } finally {
-    resetProactiveHealing();
-  }
-
-  report();
-}
-
-main();

package/dist/resources/extensions/gsd/tests/prompt-budget-enforcement.test.ts

@@ -1,464 +0,0 @@
-/**
- * Prompt budget enforcement tests — verifies that budget-aware prompt builders
- * truncate content at section boundaries and that plan-slice includes executor
- * context constraints.
- *
- * Tests:
- *   1. inlineDependencySummaries() truncates when budget is small, passes through when large
- *   2. plan-slice.md template includes {{executorContextConstraints}} placeholder
- *   3. Executor constraints formatting varies with context window size
- *   4. Different context windows produce different budget-constrained outputs
- */
-
-import { describe, it, beforeEach, afterEach } from "node:test";
-import assert from "node:assert/strict";
-import { mkdtempSync, mkdirSync, rmSync, writeFileSync, readFileSync } from "node:fs";
-import { join, dirname } from "node:path";
-import { tmpdir } from "node:os";
-import { fileURLToPath } from "node:url";
-
-import { inlineDependencySummaries } from "../auto-prompts.js";
-import { computeBudgets, truncateAtSectionBoundary } from "../context-budget.js";
-
-const __dirname = dirname(fileURLToPath(import.meta.url));
-
-// ─── Fixture helpers ──────────────────────────────────────────────────────────
-
-function createFixtureBase(): string {
-  return mkdtempSync(join(tmpdir(), "gsd-prompt-budget-test-"));
-}
-
-function cleanup(base: string): void {
-  rmSync(base, { recursive: true, force: true });
-}
-
-/**
- * Set up a minimal milestone with a roadmap declaring slice dependencies and
- * dependency slice summaries on disk.
- */
-function setupDependencyFixture(
-  base: string,
-  mid: string,
-  sid: string,
-  deps: string[],
-  summaries: Record<string, string>,
-): void {
-  const msDir = join(base, ".gsd", "milestones", mid);
-  mkdirSync(msDir, { recursive: true });
-
-  // Build roadmap content — sid depends on deps
-  const depStr = deps.join(", ");
-  const sliceLines = [
-    `- [x] **${deps[0]}: Done dep** \`risk:low\` \`depends:[]\``,
-    `- [ ] **${sid}: Current slice** \`risk:medium\` \`depends:[${depStr}]\``,
-  ];
-  // Add any extra deps as completed slices
-  for (let i = 1; i < deps.length; i++) {
-    sliceLines.unshift(`- [x] **${deps[i]}: Another dep** \`risk:low\` \`depends:[]\``);
-  }
-  const roadmapContent = [
-    "# Roadmap",
-    "",
-    "## Slices",
-    "",
-    ...sliceLines,
-  ].join("\n");
-  writeFileSync(join(msDir, `${mid}-ROADMAP.md`), roadmapContent);
-
-  // Write dependency slice summaries
-  for (const [depId, content] of Object.entries(summaries)) {
-    const sliceDir = join(msDir, "slices", depId);
-    mkdirSync(sliceDir, { recursive: true });
-    writeFileSync(join(sliceDir, `${depId}-SUMMARY.md`), content);
-  }
-
-  // Ensure target slice dir exists
-  const targetSliceDir = join(msDir, "slices", sid);
-  mkdirSync(targetSliceDir, { recursive: true });
-}
-
-// ─── inlineDependencySummaries truncation ─────────────────────────────────────
-
-describe("prompt-budget: inlineDependencySummaries truncation", () => {
-  let base: string;
-
-  beforeEach(() => {
-    base = createFixtureBase();
-  });
-
-  afterEach(() => {
-    cleanup(base);
-  });
-
-  it("passes through all content when budget is larger than total", async () => {
-    const summaryContent = "### Results\n\nEverything works.\n\n### Forward Intelligence\n\nWatch out for X.";
-    setupDependencyFixture(base, "M001", "S02", ["S01"], {
-      S01: summaryContent,
-    });
-
-    const result = await inlineDependencySummaries("M001", "S02", base, 100_000);
-    assert.ok(result.includes("Everything works."), "should include full summary content");
-    assert.ok(result.includes("Watch out for X."), "should include forward intelligence");
-    assert.ok(!result.includes("[...truncated"), "should not have truncation marker");
-  });
-
-  it("truncates at section boundaries when budget is small", async () => {
-    // Create a large summary with multiple sections
-    const sections = [];
-    for (let i = 0; i < 10; i++) {
-      sections.push(`### Section ${i}\n\n${"Lorem ipsum dolor sit amet. ".repeat(50)}`);
-    }
-    const largeSummary = sections.join("\n\n");
-
-    setupDependencyFixture(base, "M001", "S02", ["S01"], {
-      S01: largeSummary,
-    });
-
-    // Use a budget smaller than total content
-    const result = await inlineDependencySummaries("M001", "S02", base, 500);
-    assert.ok(result.includes("[...truncated"), "should have truncation marker when over budget");
-    assert.ok(result.length <= 600, `result should be near budget limit, got ${result.length}`);
-  });
-
-  it("returns content unchanged when no budget is provided (backward compat)", async () => {
-    const sections = [];
-    for (let i = 0; i < 5; i++) {
-      sections.push(`### Section ${i}\n\n${"Content block. ".repeat(30)}`);
-    }
-    const largeSummary = sections.join("\n\n");
-
-    setupDependencyFixture(base, "M001", "S02", ["S01"], {
-      S01: largeSummary,
-    });
-
-    // No budget parameter — backward-compatible behavior
-    const result = await inlineDependencySummaries("M001", "S02", base);
-    assert.ok(!result.includes("[...truncated"), "should not truncate without budget");
-    assert.ok(result.includes("Section 4"), "should include all sections");
-  });
-
-  it("handles multiple dependency summaries with truncation", async () => {
-    const summary1 = "### S01 Results\n\nFirst dep done.\n\n### S01 Notes\n\nSome notes.";
-    const summary2 = "### S02 Results\n\nSecond dep done.\n\n### S02 Notes\n\nMore notes.";
-    setupDependencyFixture(base, "M001", "S03", ["S01", "S02"], {
-      S01: summary1,
-      S02: summary2,
-    });
-
-    // Budget large enough for all content
-    const fullResult = await inlineDependencySummaries("M001", "S03", base, 100_000);
-    assert.ok(fullResult.includes("First dep done."), "should have S01 content");
-    assert.ok(fullResult.includes("Second dep done."), "should have S02 content");
-
-    // Budget too small for all
-    const truncResult = await inlineDependencySummaries("M001", "S03", base, 200);
-    assert.ok(truncResult.includes("[...truncated"), "should truncate when budget is small");
-  });
-
-  it("returns no-dependencies marker when slice has no deps", async () => {
-    const msDir = join(base, ".gsd", "milestones", "M001");
-    mkdirSync(msDir, { recursive: true });
-    const roadmap = "# Roadmap\n\n## Slices\n\n- [ ] **S01: Solo** `risk:low` `depends:[]`\n";
-    writeFileSync(join(msDir, "M001-ROADMAP.md"), roadmap);
-
-    const result = await inlineDependencySummaries("M001", "S01", base, 1000);
-    assert.equal(result, "- (no dependencies)");
-  });
-});
-
-// ─── plan-slice template includes executor constraints placeholder ────────────
-
-describe("prompt-budget: plan-slice template", () => {
-  it("contains {{executorContextConstraints}} placeholder", () => {
-    const templatePath = join(__dirname, "..", "prompts", "plan-slice.md");
-    const template = readFileSync(templatePath, "utf-8");
-    assert.ok(
-      template.includes("{{executorContextConstraints}}"),
-      "plan-slice.md should contain {{executorContextConstraints}} placeholder",
-    );
-  });
-});
-
-// ─── Executor constraints formatting ──────────────────────────────────────────
-
-describe("prompt-budget: executor constraints formatting", () => {
-  it("128K window produces different constraints than 1M window", () => {
-    const budget128K = computeBudgets(128_000);
-    const budget1M = computeBudgets(1_000_000);
-
-    // Task count ranges should differ
-    assert.notEqual(
-      budget128K.taskCountRange.max,
-      budget1M.taskCountRange.max,
-      "128K and 1M should have different max task counts",
-    );
-
-    // Inline context budgets should differ
-    assert.ok(
-      budget1M.inlineContextBudgetChars > budget128K.inlineContextBudgetChars,
-      "1M should have larger inline context budget than 128K",
-    );
-
-    // Format constraint blocks and verify they differ
-    const format = (b: ReturnType<typeof computeBudgets>, windowTokens: number) => {
-      const { min, max } = b.taskCountRange;
-      const execWindowK = Math.round(windowTokens / 1000);
-      const perTaskBudgetK = Math.round(b.inlineContextBudgetChars / 1000);
-      return [
-        `## Executor Context Constraints`,
-        ``,
-        `The agent that executes each task has a **${execWindowK}K token** context window.`,
-        `- Recommended task count for this slice: **${min}–${max} tasks**`,
-        `- Each task gets ~${perTaskBudgetK}K chars of inline context (plans, code, decisions)`,
-        `- Keep individual tasks completable within a single context window — if a task needs more context than fits, split it`,
-      ].join("\n");
-    };
-
-    const constraints128K = format(budget128K, 128_000);
-    const constraints1M = format(budget1M, 1_000_000);
-
-    assert.ok(constraints128K.includes("128K token"), "128K constraints should reference 128K");
-    assert.ok(constraints1M.includes("1000K token"), "1M constraints should reference 1000K");
-    assert.ok(constraints128K.includes("2–5 tasks"), "128K should recommend 2–5 tasks");
-    assert.ok(constraints1M.includes("2–8 tasks"), "1M should recommend 2–8 tasks");
-    assert.notEqual(constraints128K, constraints1M, "constraint blocks should differ");
-  });
-
-  it("undefined context window falls back to 200K defaults", () => {
-    // computeBudgets(0) defaults to 200K (D002)
-    const budgetDefault = computeBudgets(0);
-    const budget200K = computeBudgets(200_000);
-
-    assert.equal(budgetDefault.summaryBudgetChars, budget200K.summaryBudgetChars);
-    assert.equal(budgetDefault.inlineContextBudgetChars, budget200K.inlineContextBudgetChars);
-    assert.equal(budgetDefault.taskCountRange.max, budget200K.taskCountRange.max);
-  });
-});
-
-// ─── Budget-constrained output varies with context window ─────────────────────
-
-describe("prompt-budget: different context windows produce different outputs", () => {
-  it("small window truncates content that large window preserves", () => {
-    // Simulate assembled inlinedContext with multiple sections
-    const sections = [];
-    for (let i = 0; i < 20; i++) {
-      sections.push(`### Section ${i}: Important Context\n\n${"Detailed content for this section. ".repeat(100)}`);
-    }
-    const largeContent = `## Inlined Context\n\n${sections.join("\n\n---\n\n")}`;
-
-    // 128K context window budget
-    const budget128K = computeBudgets(128_000);
-    const r128K = truncateAtSectionBoundary(largeContent, budget128K.inlineContextBudgetChars);
-
-    // 1M context window budget
-    const budget1M = computeBudgets(1_000_000);
-    const r1M = truncateAtSectionBoundary(largeContent, budget1M.inlineContextBudgetChars);
-
-    // The large content (~70K chars) should fit in 1M budget (~1.6M chars) but
-    // if we make content bigger, the 128K budget (~204K chars) would truncate
-    assert.ok(
-      r128K.content.length <= budget128K.inlineContextBudgetChars + 100, // +100 for truncation marker
-      "128K result should respect budget",
-    );
-    assert.ok(
-      r1M.content.length <= budget1M.inlineContextBudgetChars + 100,
-      "1M result should respect budget",
-    );
-
-    // With content smaller than both budgets, both should pass through unchanged
-    const smallContent = "### One Section\n\nSmall content.";
-    const small128K = truncateAtSectionBoundary(smallContent, budget128K.inlineContextBudgetChars);
-    const small1M = truncateAtSectionBoundary(smallContent, budget1M.inlineContextBudgetChars);
-    assert.equal(small128K.content, smallContent, "small content unchanged for 128K");
-    assert.equal(small128K.droppedSections, 0);
-    assert.equal(small1M.content, smallContent, "small content unchanged for 1M");
-    assert.equal(small1M.droppedSections, 0);
-  });
-
-  it("128K budget truncates very large content while 1M preserves it", () => {
-    // Create content that exceeds 128K budget (~204K chars) but fits in 1M (~1.6M chars)
-    const sections = [];
-    for (let i = 0; i < 100; i++) {
-      sections.push(`### Section ${i}\n\n${"X".repeat(3000)}`);
-    }
-    const content = sections.join("\n\n");
-    // ~310K chars total
-
-    const budget128K = computeBudgets(128_000);
-    const result128K = truncateAtSectionBoundary(content, budget128K.inlineContextBudgetChars);
-
-    const budget1M = computeBudgets(1_000_000);
-    const result1M = truncateAtSectionBoundary(content, budget1M.inlineContextBudgetChars);
-
-    assert.ok(result128K.content.includes("[...truncated"), "128K should truncate ~310K content");
-    assert.ok(result128K.droppedSections > 0, "128K should report dropped sections");
-    assert.ok(!result1M.content.includes("[...truncated"), "1M should preserve ~310K content");
-    assert.equal(result1M.droppedSections, 0);
-    assert.ok(result128K.content.length < result1M.content.length, "128K result should be shorter than 1M result");
-  });
-});
-
-// ─── execute-task template includes verificationBudget placeholder ─────────
-
-describe("prompt-budget: execute-task template", () => {
-  it("contains {{verificationBudget}} placeholder", () => {
-    const templatePath = join(__dirname, "..", "prompts", "execute-task.md");
-    const template = readFileSync(templatePath, "utf-8");
-    assert.ok(
-      template.includes("{{verificationBudget}}"),
-      "execute-task.md should contain {{verificationBudget}} placeholder",
-    );
-  });
-
-  it("verificationBudget format varies with context window size", () => {
-    const budget128K = computeBudgets(128_000);
-    const budget1M = computeBudgets(1_000_000);
-
-    const format128K = `~${Math.round(budget128K.verificationBudgetChars / 1000)}K chars`;
-    const format1M = `~${Math.round(budget1M.verificationBudgetChars / 1000)}K chars`;
-
-    assert.notEqual(format128K, format1M, "128K and 1M should produce different verification budget strings");
-    assert.ok(format128K.includes("~51K"), `128K should produce ~51K, got ${format128K}`);
-    assert.ok(format1M.includes("~400K"), `1M should produce ~400K, got ${format1M}`);
-  });
-});
-
-// ─── buildCompleteSlicePrompt budget enforcement (simulated) ─────────────────
-
-describe("prompt-budget: complete-slice builder truncation pattern", () => {
-  it("truncateAtSectionBoundary truncates assembled inlinedContext for complete-slice pattern", () => {
-    // Simulate buildCompleteSlicePrompt: roadmap + slice plan + task summaries
-    const inlined: string[] = [];
-    inlined.push("### Milestone Roadmap\n\nRoadmap content here.");
-    inlined.push("### Slice Plan\n\nSlice plan content here.");
-    // Add many task summaries that push past budget
-    for (let i = 0; i < 50; i++) {
-      inlined.push(`### Task Summary: T${String(i).padStart(2, "0")}\nSource: \`tasks/T${String(i).padStart(2, "0")}-SUMMARY.md\`\n\n${"Task result details. ".repeat(200)}`);
-    }
-
-    const assembledContent = `## Inlined Context (preloaded — do not re-read these files)\n\n${inlined.join("\n\n---\n\n")}`;
-
-    // Small context window (128K) should truncate
-    const budget128K = computeBudgets(128_000);
-    const result128K = truncateAtSectionBoundary(assembledContent, budget128K.inlineContextBudgetChars);
-    assert.ok(result128K.content.includes("[...truncated"), "128K should truncate many task summaries");
-    assert.ok(result128K.content.includes("### Milestone Roadmap"), "should preserve early sections");
-    assert.ok(result128K.droppedSections > 0, "128K should report dropped sections");
-
-    // Large context window (1M) should preserve all
-    const budget1M = computeBudgets(1_000_000);
-    const result1M = truncateAtSectionBoundary(assembledContent, budget1M.inlineContextBudgetChars);
-    assert.ok(!result1M.content.includes("[...truncated"), "1M should preserve all task summaries");
-    assert.equal(result1M.droppedSections, 0);
-  });
-
-  it("small content passes through unchanged at any context window size", () => {
-    const smallContent = "## Inlined Context\n\n### Roadmap\n\nSmall roadmap.\n\n---\n\n### Plan\n\nSmall plan.";
-
-    const budget128K = computeBudgets(128_000);
-    const result128K = truncateAtSectionBoundary(smallContent, budget128K.inlineContextBudgetChars);
-    assert.equal(result128K.content, smallContent, "small content unchanged for 128K");
-    assert.equal(result128K.droppedSections, 0);
-
-    const budget1M = computeBudgets(1_000_000);
-    const result1M = truncateAtSectionBoundary(smallContent, budget1M.inlineContextBudgetChars);
-    assert.equal(result1M.content, smallContent, "small content unchanged for 1M");
-    assert.equal(result1M.droppedSections, 0);
-  });
-});
-
-// ─── buildCompleteMilestonePrompt budget enforcement (simulated) ─────────────
-
-describe("prompt-budget: complete-milestone builder truncation pattern", () => {
-  it("truncateAtSectionBoundary truncates assembled inlinedContext for complete-milestone pattern", () => {
-    // Simulate buildCompleteMilestonePrompt: roadmap + slice summaries + root files
-    const inlined: string[] = [];
-    inlined.push("### Milestone Roadmap\n\nRoadmap content here.");
-    // Add many slice summaries that push past budget
-    for (let i = 0; i < 30; i++) {
-      inlined.push(`### S${String(i).padStart(2, "0")} Summary\n\n${"Slice summary with detailed results and forward intelligence. ".repeat(200)}`);
-    }
-    inlined.push("### Requirements\n\nProject requirements.");
-    inlined.push("### Decisions\n\nProject decisions.");
-
-    const assembledContent = `## Inlined Context (preloaded — do not re-read these files)\n\n${inlined.join("\n\n---\n\n")}`;
-
-    // Small context window (128K) should truncate
-    const budget128K = computeBudgets(128_000);
-    const result128K = truncateAtSectionBoundary(assembledContent, budget128K.inlineContextBudgetChars);
-    assert.ok(result128K.content.includes("[...truncated"), "128K should truncate many slice summaries");
-    assert.ok(result128K.droppedSections > 0);
-
-    // Large context window (1M) should preserve all
-    const budget1M = computeBudgets(1_000_000);
-    const result1M = truncateAtSectionBoundary(assembledContent, budget1M.inlineContextBudgetChars);
-    assert.ok(!result1M.content.includes("[...truncated"), "1M should preserve all slice summaries");
-    assert.equal(result1M.droppedSections, 0);
-  });
-
-  it("different context windows produce different truncation for milestone completion", () => {
-    // Create content that exceeds 128K budget but not 200K budget
-    const inlined: string[] = [];
-    inlined.push("### Roadmap\n\nRoadmap.");
-    for (let i = 0; i < 15; i++) {
-      inlined.push(`### S${i} Summary\n\n${"X".repeat(15000)}`);
-    }
-    const content = `## Inlined Context\n\n${inlined.join("\n\n---\n\n")}`;
-    // ~225K chars total
-
-    const budget128K = computeBudgets(128_000);
-    const budget200K = computeBudgets(200_000);
-    const budget1M = computeBudgets(1_000_000);
-
-    const result128K = truncateAtSectionBoundary(content, budget128K.inlineContextBudgetChars);
-    const result200K = truncateAtSectionBoundary(content, budget200K.inlineContextBudgetChars);
-    const result1M = truncateAtSectionBoundary(content, budget1M.inlineContextBudgetChars);
-
-    // 128K (budget ~204K) should truncate ~225K content
-    assert.ok(result128K.content.includes("[...truncated"), "128K should truncate ~225K content");
-    assert.ok(result128K.droppedSections > 0);
-    // 200K (budget ~320K) should not truncate ~225K content
-    assert.ok(!result200K.content.includes("[...truncated"), "200K should preserve ~225K content");
-    assert.equal(result200K.droppedSections, 0);
-    // 1M should not truncate
-    assert.ok(!result1M.content.includes("[...truncated"), "1M should preserve ~225K content");
-    assert.equal(result1M.droppedSections, 0);
-    // 128K result should be shorter
-    assert.ok(result128K.content.length < result200K.content.length, "128K result should be shorter than 200K");
-  });
-});
-
-// ─── buildExecuteTaskPrompt budget enforcement (simulated) ───────────────────
-
-describe("prompt-budget: execute-task builder truncation pattern", () => {
-  it("truncateAtSectionBoundary truncates assembled carry-forward + task plan + slice excerpt", () => {
-    // Simulate the assembled content from buildExecuteTaskPrompt
-    const carryForward = "## Carry-Forward Context\n" + Array.from({ length: 20 }, (_, i) =>
-      `- \`tasks/T${String(i).padStart(2, "0")}-SUMMARY.md\` — ${"Summary details. ".repeat(100)}`
-    ).join("\n");
-
-    const taskPlan = "## Inlined Task Plan\n\n" + Array.from({ length: 10 }, (_, i) =>
-      `### Step ${i}\n\n${"Implementation step details. ".repeat(200)}`
-    ).join("\n\n");
-
-    const sliceExcerpt = "## Slice Plan Excerpt\n\n" + "Slice goal and verification details. ".repeat(100);
-
-    const assembled = [carryForward, taskPlan, sliceExcerpt].join("\n\n---\n\n");
-
-    // Small context window should truncate
-    const budget128K = computeBudgets(128_000);
-    const result = truncateAtSectionBoundary(assembled, budget128K.inlineContextBudgetChars);
-
-    // Content should respect budget
-    assert.ok(
-      result.content.length <= budget128K.inlineContextBudgetChars + 100,
-      `result should respect 128K budget, got ${result.content.length} chars vs budget ${budget128K.inlineContextBudgetChars}`,
-    );
-
-    // Large content should be truncated
-    if (assembled.length > budget128K.inlineContextBudgetChars) {
-      assert.ok(result.content.includes("[...truncated"), "should truncate when content exceeds 128K budget");
-      assert.ok(result.droppedSections > 0, "should report dropped sections");
-    }
-  });
-});