@izkac/forgekit 0.3.12 → 0.3.13

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (116) hide show
  1. package/bin/forge.mjs +107 -107
  2. package/bin/forgekit.mjs +83 -83
  3. package/bin/review.mjs +81 -81
  4. package/package.json +1 -1
  5. package/scripts/prepack.mjs +78 -78
  6. package/scripts/run-tests.mjs +50 -50
  7. package/src/adr.mjs +236 -236
  8. package/src/adr.test.mjs +170 -170
  9. package/src/change.mjs +327 -234
  10. package/src/change.test.mjs +145 -83
  11. package/src/cleanup-sessions.mjs +84 -84
  12. package/src/config.mjs +103 -103
  13. package/src/defer.mjs +75 -75
  14. package/src/doctor.mjs +350 -341
  15. package/src/doctor.test.mjs +114 -114
  16. package/src/init.mjs +680 -621
  17. package/src/install.mjs +815 -815
  18. package/src/install.test.mjs +180 -180
  19. package/src/integrity-check.mjs +60 -60
  20. package/src/integrity.mjs +682 -682
  21. package/src/integrity.test.mjs +566 -566
  22. package/src/lib.mjs +143 -143
  23. package/src/models.defaults.json +41 -41
  24. package/src/new-session.mjs +99 -99
  25. package/src/openspec-overlays/README.md +19 -19
  26. package/src/openspec-overlays/openspec-apply-change-footer.md +14 -14
  27. package/src/openspec-overlays/opsx-apply-completion-step.md +1 -1
  28. package/src/openspec-overlays/opsx-apply-implement-step.md +11 -11
  29. package/src/paths.mjs +92 -92
  30. package/src/plan-engine.mjs +321 -278
  31. package/src/plan-engine.test.mjs +447 -283
  32. package/src/preferences.defaults.json +78 -78
  33. package/src/preferences.mjs +438 -438
  34. package/src/preferences.test.mjs +174 -174
  35. package/src/record-evidence.mjs +204 -204
  36. package/src/resolve-model.mjs +312 -312
  37. package/src/resolve-model.test.mjs +194 -194
  38. package/src/review/cli.test.mjs +117 -117
  39. package/src/review/export.mjs +172 -172
  40. package/src/review/export.test.mjs +197 -197
  41. package/src/review/fixtures/valid-review.json +42 -42
  42. package/src/review/lib.mjs +894 -894
  43. package/src/review/lib.test.mjs +266 -266
  44. package/src/review/schema.json +196 -196
  45. package/src/review/signals.test.mjs +62 -62
  46. package/src/score-cli.mjs +68 -68
  47. package/src/score.mjs +568 -568
  48. package/src/score.test.mjs +366 -366
  49. package/src/session-reminder.mjs +207 -207
  50. package/src/session-status.mjs +70 -70
  51. package/src/set-models.mjs +186 -186
  52. package/src/set-phase.mjs +205 -205
  53. package/src/set-prefs.mjs +294 -294
  54. package/src/specs-sync.mjs +234 -0
  55. package/src/specs-sync.test.mjs +114 -0
  56. package/src/spine.mjs +93 -93
  57. package/src/triage-prompt.mjs +175 -175
  58. package/src/triage-prompt.test.mjs +50 -50
  59. package/src/vendor-openspec-overlays.mjs +176 -176
  60. package/src/vendor-openspec-overlays.test.mjs +62 -62
  61. package/vendor/skills/archive-to-adr/SKILL.md +149 -149
  62. package/vendor/skills/forge/SKILL.md +136 -136
  63. package/vendor/skills/forge/docs/forge.md +650 -647
  64. package/vendor/skills/forge/phases/brainstorm.md +23 -23
  65. package/vendor/skills/forge/phases/finish.md +90 -87
  66. package/vendor/skills/forge/phases/implement.md +77 -77
  67. package/vendor/skills/forge/phases/plan-openspec.md +60 -60
  68. package/vendor/skills/forge/phases/plan-specs.md +163 -117
  69. package/vendor/skills/forge/phases/review.md +25 -25
  70. package/vendor/skills/forge/phases/verify.md +124 -124
  71. package/vendor/skills/forge/references/forge-layout.md +85 -85
  72. package/vendor/skills/forge/references/pace.md +115 -115
  73. package/vendor/skills/forge/references/plan-routing.md +52 -51
  74. package/vendor/skills/forge/references/runtime-integrity.md +232 -232
  75. package/vendor/skills/forge/references/substantial-work.md +37 -37
  76. package/vendor/skills/forge/references/test-evidence.md +30 -30
  77. package/vendor/skills/forge/references/test-strategy.md +68 -68
  78. package/vendor/skills/forge/skills/subagent-driven-development/SKILL.md +87 -87
  79. package/vendor/skills/forge/subagents/final-reviewer-prompt.md +56 -56
  80. package/vendor/skills/forge/subagents/implementer-prompt.md +38 -38
  81. package/vendor/skills/git-resolve-adr-conflict/SKILL.md +132 -132
  82. package/vendor/skills/thorough-code-review/SKILL.md +290 -290
  83. package/vendor/skills/thorough-code-review/examples.md +133 -133
  84. package/vendor/skills/thorough-code-review/reference/accepted-risks.md +26 -26
  85. package/vendor/skills/thorough-code-review/reference/lenses.md +96 -96
  86. package/vendor/skills/thorough-code-review/reference/phase1-scout.md +62 -62
  87. package/vendor/skills/thorough-code-review/reference/phase2-skeptic.md +105 -105
  88. package/vendor/skills/thorough-code-review/reference/report-schema.json +222 -222
  89. package/vendor/skills/thorough-code-review/reference/report-template.md +115 -115
  90. package/vendor/skills/thorough-code-review/reference/severity-rubric.md +49 -49
  91. package/vendor/skills/thorough-code-review/reference/signals-preflight.md +55 -55
  92. package/vendor/templates/adr/README.md +7 -7
  93. package/vendor/templates/adr/decisions.md +141 -141
  94. package/vendor/templates/adr/hooks/check-pending-adrs.mjs +74 -74
  95. package/vendor/templates/adr/hooks/check-pending-adrs.sh +3 -3
  96. package/vendor/templates/adr/hooks/openspec-archive-agent-message.mjs +52 -52
  97. package/vendor/templates/adr/hooks/openspec-archive-agent-message.sh +3 -3
  98. package/vendor/templates/project/claude/commands/forge-apply.md +75 -75
  99. package/vendor/templates/project/claude/commands/forge-brainstorm.md +7 -7
  100. package/vendor/templates/project/claude/commands/forge-build.md +17 -17
  101. package/vendor/templates/project/claude/commands/forge-plan.md +12 -12
  102. package/vendor/templates/project/claude/commands/forge-skip.md +14 -14
  103. package/vendor/templates/project/claude/commands/forge-status.md +16 -16
  104. package/vendor/templates/project/claude/commands/forge.md +16 -16
  105. package/vendor/templates/project/claude/hooks/forge-prompt-hook.mjs +73 -73
  106. package/vendor/templates/project/claude/hooks/forge-session-start.mjs +19 -19
  107. package/vendor/templates/project/claude/hooks/forge-triage-hook.mjs +77 -77
  108. package/vendor/templates/project/cursor/commands/forge-apply.md +75 -75
  109. package/vendor/templates/project/cursor/commands/forge-brainstorm.md +10 -10
  110. package/vendor/templates/project/cursor/commands/forge-build.md +17 -17
  111. package/vendor/templates/project/cursor/commands/forge-plan.md +15 -15
  112. package/vendor/templates/project/cursor/commands/forge-skip.md +14 -14
  113. package/vendor/templates/project/cursor/commands/forge-status.md +16 -16
  114. package/vendor/templates/project/cursor/commands/forge.md +16 -16
  115. package/vendor/templates/project/cursor/hooks/forge-session-start.mjs +30 -30
  116. package/vendor/templates/project/cursor/hooks/forge-session-start.sh +3 -3
@@ -1,366 +1,366 @@
1
- import test from 'node:test';
2
- import assert from 'node:assert/strict';
3
- import fs from 'node:fs';
4
- import path from 'node:path';
5
- import { tmpdir } from 'node:os';
6
- import { execFileSync } from 'node:child_process';
7
- import { fileURLToPath } from 'node:url';
8
- import {
9
- formatScorecardMarkdown,
10
- gradeForScore,
11
- scoreSession,
12
- writeSessionScorecard,
13
- } from './score.mjs';
14
- import { e2eStepsHash } from './integrity.mjs';
15
-
16
- const PHASE_SCRIPT = path.join(path.dirname(fileURLToPath(import.meta.url)), 'set-phase.mjs');
17
- const SCORE_SCRIPT = path.join(path.dirname(fileURLToPath(import.meta.url)), 'score-cli.mjs');
18
- const CLEANUP_SCRIPT = path.join(path.dirname(fileURLToPath(import.meta.url)), 'cleanup-sessions.mjs');
19
-
20
- function tmp(prefix) {
21
- return fs.mkdtempSync(path.join(tmpdir(), prefix));
22
- }
23
-
24
- function makeSession(root, overrides = {}) {
25
- const sessionDir = path.join(root, '.forge', 'sessions', 'sess-score');
26
- fs.mkdirSync(sessionDir, { recursive: true });
27
- const now = new Date().toISOString();
28
- const session = {
29
- id: 'sess-score',
30
- slug: 'fixture',
31
- createdAt: now,
32
- updatedAt: now,
33
- phase: 'verify',
34
- planType: null,
35
- openspecChange: null,
36
- tasksTotal: 2,
37
- tasksComplete: 2,
38
- pace: 'auto',
39
- resolvedPace: 'standard',
40
- pacePinned: false,
41
- ...overrides,
42
- };
43
- fs.writeFileSync(path.join(sessionDir, 'session.json'), `${JSON.stringify(session, null, 2)}\n`, 'utf8');
44
- fs.writeFileSync(
45
- path.join(root, '.forge', 'active.json'),
46
- `${JSON.stringify({ sessionId: 'sess-score' }, null, 2)}\n`,
47
- 'utf8',
48
- );
49
- return { sessionDir, session };
50
- }
51
-
52
- function validSpine(rows) {
53
- return {
54
- change: null,
55
- notApplicable: null,
56
- rows,
57
- };
58
- }
59
-
60
- function validRow() {
61
- return {
62
- capability: 'REQ-GOV-01 matching',
63
- library: 'etl_core/matcher.py',
64
- runtimeOwner: 'worker job analyze_study',
65
- writes: 'study_proposals',
66
- reads: 'N/A',
67
- uiConsumer: 'Proposals',
68
- evidence: 'tasks/12/test-evidence.md',
69
- };
70
- }
71
-
72
- test('gradeForScore bands', () => {
73
- assert.equal(gradeForScore(95), 'A');
74
- assert.equal(gradeForScore(80), 'B');
75
- assert.equal(gradeForScore(60), 'C');
76
- assert.equal(gradeForScore(40), 'D');
77
- assert.equal(gradeForScore(10), 'F');
78
- });
79
-
80
- test('scoreSession: strong sync-only session scores high', () => {
81
- const root = tmp('forge-score-strong-');
82
- try {
83
- const { sessionDir, session } = makeSession(root, { slug: 'add-health' });
84
- fs.writeFileSync(
85
- path.join(sessionDir, 'spine.json'),
86
- `${JSON.stringify({ rows: [], notApplicable: 'sync HTTP only' }, null, 2)}\n`,
87
- 'utf8',
88
- );
89
- fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# Verify\n\nExit 0\n', 'utf8');
90
- const taskDir = path.join(sessionDir, 'tasks', '01-health');
91
- fs.mkdirSync(taskDir, { recursive: true });
92
- fs.writeFileSync(
93
- path.join(taskDir, 'test-evidence.md'),
94
- '# Test evidence\n\n- **Exit code:** 0\n- **Summary:** assert response.ok === true\n',
95
- 'utf8',
96
- );
97
-
98
- const card = scoreSession({ cwd: root, sessionDir, session });
99
- assert.ok(card.score >= 80, `expected >=80, got ${card.score}`);
100
- assert.ok(['A', 'B'].includes(card.grade));
101
- assert.equal(card.integrityOk, true);
102
- assert.match(formatScorecardMarkdown(card), /Grade: [AB]/);
103
- } finally {
104
- fs.rmSync(root, { recursive: true, force: true });
105
- }
106
- });
107
-
108
- test('scoreSession: missing spine scores poorly', () => {
109
- const root = tmp('forge-score-weak-');
110
- try {
111
- const { sessionDir, session } = makeSession(root);
112
- fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# ok\n', 'utf8');
113
- const card = scoreSession({ cwd: root, sessionDir, session });
114
- assert.ok(card.score < 60, `expected weak score, got ${card.score}`);
115
- assert.equal(card.integrityOk, false);
116
- } finally {
117
- fs.rmSync(root, { recursive: true, force: true });
118
- }
119
- });
120
-
121
- test('scoreSession: incompleteReason caps score at 59', () => {
122
- const root = tmp('forge-score-cap-');
123
- try {
124
- const { sessionDir, session } = makeSession(root, {
125
- incompleteReason: 'E2E blocked',
126
- slug: 'add-health',
127
- });
128
- fs.writeFileSync(
129
- path.join(sessionDir, 'spine.json'),
130
- `${JSON.stringify({ rows: [], notApplicable: 'sync HTTP only' }, null, 2)}\n`,
131
- 'utf8',
132
- );
133
- fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# ok\n', 'utf8');
134
- const card = scoreSession({ cwd: root, sessionDir, session });
135
- assert.ok(card.score <= 59);
136
- assert.ok(card.caps.some((c) => /incompleteReason/.test(c)));
137
- } finally {
138
- fs.rmSync(root, { recursive: true, force: true });
139
- }
140
- });
141
-
142
- test('scoreSession: product-loop with baseline-diff scores higher than bare section', () => {
143
- const root = tmp('forge-score-loop-');
144
- try {
145
- const { sessionDir, session } = makeSession(root, { slug: 'wire-worker-jobs' });
146
- fs.writeFileSync(
147
- path.join(sessionDir, 'spine.json'),
148
- `${JSON.stringify(validSpine([validRow()]), null, 2)}\n`,
149
- 'utf8',
150
- );
151
- fs.writeFileSync(
152
- path.join(sessionDir, 'verify-evidence.md'),
153
- '# Verify\n\n## Product loop\n\nJust a heading\n',
154
- 'utf8',
155
- );
156
- const weak = scoreSession({ cwd: root, sessionDir, session });
157
-
158
- fs.writeFileSync(
159
- path.join(sessionDir, 'verify-evidence.md'),
160
- `# Verify
161
-
162
- ## Product loop
163
-
164
- Fixture: OP1086
165
-
166
- 1. ingest
167
- 2. analyze
168
- 3. ratify
169
- 4. run @R — output differs from baseline
170
- `,
171
- 'utf8',
172
- );
173
- const strong = scoreSession({ cwd: root, sessionDir, session });
174
- assert.ok(strong.score > weak.score, `${strong.score} should beat ${weak.score}`);
175
- const loop = strong.checks.find((c) => c.id === 'product_loop');
176
- assert.ok(loop.points >= 15);
177
- } finally {
178
- fs.rmSync(root, { recursive: true, force: true });
179
- }
180
- });
181
-
182
- test('scoreSession: product-loop bonuses match inflected forms (asserts, fixtures, ratify)', () => {
183
- const root = tmp('forge-score-inflect-');
184
- try {
185
- const { sessionDir, session } = makeSession(root, { slug: 'jobs-loop' });
186
- fs.writeFileSync(
187
- path.join(sessionDir, 'spine.json'),
188
- `${JSON.stringify(validSpine([validRow()]), null, 2)}\n`,
189
- 'utf8',
190
- );
191
- // Every keyword deliberately inflected — a trailing \b used to reject all of these.
192
- fs.writeFileSync(
193
- path.join(sessionDir, 'verify-evidence.md'),
194
- '# Verify\n\n## Product loop\n\n1. seed fixtures\n2. run the loop\n3. asserts the ratify output differs from stored baselines\n',
195
- 'utf8',
196
- );
197
- const card = scoreSession({ cwd: root, sessionDir, session });
198
- const loop = card.checks.find((c) => c.id === 'product_loop');
199
- const notes = loop.notes.join('\n');
200
- assert.match(notes, /names a fixture/);
201
- assert.match(notes, /asserts decision\/output change/);
202
- assert.equal(loop.points, 20);
203
- } finally {
204
- fs.rmSync(root, { recursive: true, force: true });
205
- }
206
- });
207
-
208
- test('scoreSession: green e2e run proves the loop even without the phrase', () => {
209
- const root = tmp('forge-score-e2e-');
210
- try {
211
- const { sessionDir, session } = makeSession(root, { slug: 'jobs-loop' });
212
- fs.writeFileSync(
213
- path.join(sessionDir, 'spine.json'),
214
- `${JSON.stringify(validSpine([validRow()]), null, 2)}\n`,
215
- 'utf8',
216
- );
217
- // Evidence deliberately avoids the words "product loop" — before the fix
218
- // this session scored 0/20 despite an executed green run.
219
- fs.writeFileSync(
220
- path.join(sessionDir, 'verify-evidence.md'),
221
- '# Verify\n\n## What the evidence actually proves\n\nReverted the fix, reproduced the broken row, re-applied, ran the loop green.\n',
222
- 'utf8',
223
- );
224
- const steps = [{ name: 'probe', cmd: 'node -e "console.log(1)"' }];
225
- fs.writeFileSync(
226
- path.join(sessionDir, 'e2e.json'),
227
- `${JSON.stringify({ change: null, notApplicable: null, steps }, null, 2)}\n`,
228
- 'utf8',
229
- );
230
- fs.writeFileSync(
231
- path.join(sessionDir, 'e2e-results.json'),
232
- `${JSON.stringify({ ok: true, stepsHash: e2eStepsHash(steps), steps: [{ name: 'probe', ok: true, exitCode: 0 }] }, null, 2)}\n`,
233
- 'utf8',
234
- );
235
- const card = scoreSession({ cwd: root, sessionDir, session });
236
- const loop = card.checks.find((c) => c.id === 'product_loop');
237
- assert.ok(loop.points >= 8, `expected >= 8 loop points, got ${loop.points}`);
238
- assert.match(loop.notes.join('\n'), /proven by execution/);
239
- } finally {
240
- fs.rmSync(root, { recursive: true, force: true });
241
- }
242
- });
243
-
244
- test('forge cleanup harvests scorecard.json into the ledger before pruning', () => {
245
- const root = tmp('forge-cleanup-harvest-');
246
- try {
247
- const { sessionDir, session } = makeSession(root, { phase: 'done' });
248
- // A scorecard as an older (pre-ledger) forgekit would have written it.
249
- fs.writeFileSync(
250
- path.join(sessionDir, 'scorecard.json'),
251
- `${JSON.stringify({
252
- scoredAt: '2026-06-01T00:00:00.000Z',
253
- sessionId: session.id,
254
- slug: session.slug,
255
- openspecChange: null,
256
- score: 83,
257
- maxScore: 100,
258
- grade: 'B',
259
- integrityOk: true,
260
- caps: [],
261
- checks: [{ id: 'spine', label: 'Spine matrix quality', points: 20, max: 25, notes: ['one row thin'] }],
262
- }, null, 2)}\n`,
263
- 'utf8',
264
- );
265
- execFileSync(process.execPath, [CLEANUP_SCRIPT, '--include-active'], {
266
- cwd: root,
267
- env: { ...process.env, FORGEKIT_FLEET_DIR: path.join(root, 'fleet') },
268
- });
269
- assert.equal(fs.existsSync(sessionDir), false);
270
- const ledger = path.join(root, '.forge', 'scorecards.jsonl');
271
- const lines = fs.readFileSync(ledger, 'utf8').split('\n').filter(Boolean);
272
- assert.equal(lines.length, 1);
273
- const entry = JSON.parse(lines[0]);
274
- assert.equal(entry.sessionId, session.id);
275
- assert.equal(entry.grade, 'B');
276
- assert.equal(entry.deductions[0].id, 'spine');
277
- } finally {
278
- fs.rmSync(root, { recursive: true, force: true });
279
- }
280
- });
281
-
282
- test('writeSessionScorecard writes json and md', () => {
283
- const root = tmp('forge-score-write-');
284
- try {
285
- const { sessionDir, session } = makeSession(root, { slug: 'docs-only' });
286
- fs.writeFileSync(
287
- path.join(sessionDir, 'spine.json'),
288
- `${JSON.stringify({ rows: [], notApplicable: 'docs-only' }, null, 2)}\n`,
289
- 'utf8',
290
- );
291
- fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# ok\n', 'utf8');
292
- const { jsonPath, mdPath, card } = writeSessionScorecard({ cwd: root, sessionDir, session });
293
- assert.equal(fs.existsSync(jsonPath), true);
294
- assert.equal(fs.existsSync(mdPath), true);
295
- assert.equal(JSON.parse(fs.readFileSync(jsonPath, 'utf8')).grade, card.grade);
296
-
297
- // Durable ledger: one line per session, re-scoring replaces (not appends).
298
- const ledger = path.join(root, '.forge', 'scorecards.jsonl');
299
- assert.equal(fs.existsSync(ledger), true);
300
- writeSessionScorecard({ cwd: root, sessionDir, session });
301
- const lines = fs.readFileSync(ledger, 'utf8').split('\n').filter(Boolean);
302
- assert.equal(lines.length, 1);
303
- const entry = JSON.parse(lines[0]);
304
- assert.equal(entry.sessionId, session.id);
305
- assert.equal(entry.grade, card.grade);
306
- assert.ok(Array.isArray(entry.deductions));
307
- } finally {
308
- fs.rmSync(root, { recursive: true, force: true });
309
- }
310
- });
311
-
312
- test('forge phase done writes scorecard and stamps session.scoreGrade', () => {
313
- const root = tmp('forge-score-phase-');
314
- try {
315
- const { sessionDir } = makeSession(root, {
316
- slug: 'add-health',
317
- phase: 'verify',
318
- tasksTotal: 1,
319
- tasksComplete: 1,
320
- });
321
- fs.writeFileSync(
322
- path.join(sessionDir, 'spine.json'),
323
- `${JSON.stringify({ rows: [], notApplicable: 'sync HTTP only' }, null, 2)}\n`,
324
- 'utf8',
325
- );
326
- fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# ok\n', 'utf8');
327
-
328
- execFileSync(process.execPath, [PHASE_SCRIPT, 'done'], {
329
- cwd: root,
330
- encoding: 'utf8',
331
- stdio: ['ignore', 'pipe', 'pipe'],
332
- });
333
-
334
- assert.equal(fs.existsSync(path.join(sessionDir, 'scorecard.md')), true);
335
- assert.equal(fs.existsSync(path.join(sessionDir, 'scorecard.json')), true);
336
- const session = JSON.parse(fs.readFileSync(path.join(sessionDir, 'session.json'), 'utf8'));
337
- assert.equal(session.phase, 'done');
338
- assert.ok(typeof session.score === 'number');
339
- assert.ok(['A', 'B', 'C', 'D', 'F'].includes(session.scoreGrade));
340
- } finally {
341
- fs.rmSync(root, { recursive: true, force: true });
342
- }
343
- });
344
-
345
- test('forge score CLI prints JSON', () => {
346
- const root = tmp('forge-score-cli-');
347
- try {
348
- const { sessionDir } = makeSession(root, { slug: 'add-health' });
349
- fs.writeFileSync(
350
- path.join(sessionDir, 'spine.json'),
351
- `${JSON.stringify({ rows: [], notApplicable: 'sync HTTP only' }, null, 2)}\n`,
352
- 'utf8',
353
- );
354
- fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# ok\n', 'utf8');
355
- const out = execFileSync(process.execPath, [SCORE_SCRIPT], {
356
- cwd: root,
357
- encoding: 'utf8',
358
- });
359
- const card = JSON.parse(out);
360
- assert.ok(card.grade);
361
- assert.ok(Array.isArray(card.humanPrompts));
362
- assert.equal(card.humanPrompts.length, 4);
363
- } finally {
364
- fs.rmSync(root, { recursive: true, force: true });
365
- }
366
- });
1
+ import test from 'node:test';
2
+ import assert from 'node:assert/strict';
3
+ import fs from 'node:fs';
4
+ import path from 'node:path';
5
+ import { tmpdir } from 'node:os';
6
+ import { execFileSync } from 'node:child_process';
7
+ import { fileURLToPath } from 'node:url';
8
+ import {
9
+ formatScorecardMarkdown,
10
+ gradeForScore,
11
+ scoreSession,
12
+ writeSessionScorecard,
13
+ } from './score.mjs';
14
+ import { e2eStepsHash } from './integrity.mjs';
15
+
16
+ const PHASE_SCRIPT = path.join(path.dirname(fileURLToPath(import.meta.url)), 'set-phase.mjs');
17
+ const SCORE_SCRIPT = path.join(path.dirname(fileURLToPath(import.meta.url)), 'score-cli.mjs');
18
+ const CLEANUP_SCRIPT = path.join(path.dirname(fileURLToPath(import.meta.url)), 'cleanup-sessions.mjs');
19
+
20
+ function tmp(prefix) {
21
+ return fs.mkdtempSync(path.join(tmpdir(), prefix));
22
+ }
23
+
24
+ function makeSession(root, overrides = {}) {
25
+ const sessionDir = path.join(root, '.forge', 'sessions', 'sess-score');
26
+ fs.mkdirSync(sessionDir, { recursive: true });
27
+ const now = new Date().toISOString();
28
+ const session = {
29
+ id: 'sess-score',
30
+ slug: 'fixture',
31
+ createdAt: now,
32
+ updatedAt: now,
33
+ phase: 'verify',
34
+ planType: null,
35
+ openspecChange: null,
36
+ tasksTotal: 2,
37
+ tasksComplete: 2,
38
+ pace: 'auto',
39
+ resolvedPace: 'standard',
40
+ pacePinned: false,
41
+ ...overrides,
42
+ };
43
+ fs.writeFileSync(path.join(sessionDir, 'session.json'), `${JSON.stringify(session, null, 2)}\n`, 'utf8');
44
+ fs.writeFileSync(
45
+ path.join(root, '.forge', 'active.json'),
46
+ `${JSON.stringify({ sessionId: 'sess-score' }, null, 2)}\n`,
47
+ 'utf8',
48
+ );
49
+ return { sessionDir, session };
50
+ }
51
+
52
+ function validSpine(rows) {
53
+ return {
54
+ change: null,
55
+ notApplicable: null,
56
+ rows,
57
+ };
58
+ }
59
+
60
+ function validRow() {
61
+ return {
62
+ capability: 'REQ-GOV-01 matching',
63
+ library: 'etl_core/matcher.py',
64
+ runtimeOwner: 'worker job analyze_study',
65
+ writes: 'study_proposals',
66
+ reads: 'N/A',
67
+ uiConsumer: 'Proposals',
68
+ evidence: 'tasks/12/test-evidence.md',
69
+ };
70
+ }
71
+
72
+ test('gradeForScore bands', () => {
73
+ assert.equal(gradeForScore(95), 'A');
74
+ assert.equal(gradeForScore(80), 'B');
75
+ assert.equal(gradeForScore(60), 'C');
76
+ assert.equal(gradeForScore(40), 'D');
77
+ assert.equal(gradeForScore(10), 'F');
78
+ });
79
+
80
+ test('scoreSession: strong sync-only session scores high', () => {
81
+ const root = tmp('forge-score-strong-');
82
+ try {
83
+ const { sessionDir, session } = makeSession(root, { slug: 'add-health' });
84
+ fs.writeFileSync(
85
+ path.join(sessionDir, 'spine.json'),
86
+ `${JSON.stringify({ rows: [], notApplicable: 'sync HTTP only' }, null, 2)}\n`,
87
+ 'utf8',
88
+ );
89
+ fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# Verify\n\nExit 0\n', 'utf8');
90
+ const taskDir = path.join(sessionDir, 'tasks', '01-health');
91
+ fs.mkdirSync(taskDir, { recursive: true });
92
+ fs.writeFileSync(
93
+ path.join(taskDir, 'test-evidence.md'),
94
+ '# Test evidence\n\n- **Exit code:** 0\n- **Summary:** assert response.ok === true\n',
95
+ 'utf8',
96
+ );
97
+
98
+ const card = scoreSession({ cwd: root, sessionDir, session });
99
+ assert.ok(card.score >= 80, `expected >=80, got ${card.score}`);
100
+ assert.ok(['A', 'B'].includes(card.grade));
101
+ assert.equal(card.integrityOk, true);
102
+ assert.match(formatScorecardMarkdown(card), /Grade: [AB]/);
103
+ } finally {
104
+ fs.rmSync(root, { recursive: true, force: true });
105
+ }
106
+ });
107
+
108
+ test('scoreSession: missing spine scores poorly', () => {
109
+ const root = tmp('forge-score-weak-');
110
+ try {
111
+ const { sessionDir, session } = makeSession(root);
112
+ fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# ok\n', 'utf8');
113
+ const card = scoreSession({ cwd: root, sessionDir, session });
114
+ assert.ok(card.score < 60, `expected weak score, got ${card.score}`);
115
+ assert.equal(card.integrityOk, false);
116
+ } finally {
117
+ fs.rmSync(root, { recursive: true, force: true });
118
+ }
119
+ });
120
+
121
+ test('scoreSession: incompleteReason caps score at 59', () => {
122
+ const root = tmp('forge-score-cap-');
123
+ try {
124
+ const { sessionDir, session } = makeSession(root, {
125
+ incompleteReason: 'E2E blocked',
126
+ slug: 'add-health',
127
+ });
128
+ fs.writeFileSync(
129
+ path.join(sessionDir, 'spine.json'),
130
+ `${JSON.stringify({ rows: [], notApplicable: 'sync HTTP only' }, null, 2)}\n`,
131
+ 'utf8',
132
+ );
133
+ fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# ok\n', 'utf8');
134
+ const card = scoreSession({ cwd: root, sessionDir, session });
135
+ assert.ok(card.score <= 59);
136
+ assert.ok(card.caps.some((c) => /incompleteReason/.test(c)));
137
+ } finally {
138
+ fs.rmSync(root, { recursive: true, force: true });
139
+ }
140
+ });
141
+
142
+ test('scoreSession: product-loop with baseline-diff scores higher than bare section', () => {
143
+ const root = tmp('forge-score-loop-');
144
+ try {
145
+ const { sessionDir, session } = makeSession(root, { slug: 'wire-worker-jobs' });
146
+ fs.writeFileSync(
147
+ path.join(sessionDir, 'spine.json'),
148
+ `${JSON.stringify(validSpine([validRow()]), null, 2)}\n`,
149
+ 'utf8',
150
+ );
151
+ fs.writeFileSync(
152
+ path.join(sessionDir, 'verify-evidence.md'),
153
+ '# Verify\n\n## Product loop\n\nJust a heading\n',
154
+ 'utf8',
155
+ );
156
+ const weak = scoreSession({ cwd: root, sessionDir, session });
157
+
158
+ fs.writeFileSync(
159
+ path.join(sessionDir, 'verify-evidence.md'),
160
+ `# Verify
161
+
162
+ ## Product loop
163
+
164
+ Fixture: OP1086
165
+
166
+ 1. ingest
167
+ 2. analyze
168
+ 3. ratify
169
+ 4. run @R — output differs from baseline
170
+ `,
171
+ 'utf8',
172
+ );
173
+ const strong = scoreSession({ cwd: root, sessionDir, session });
174
+ assert.ok(strong.score > weak.score, `${strong.score} should beat ${weak.score}`);
175
+ const loop = strong.checks.find((c) => c.id === 'product_loop');
176
+ assert.ok(loop.points >= 15);
177
+ } finally {
178
+ fs.rmSync(root, { recursive: true, force: true });
179
+ }
180
+ });
181
+
182
+ test('scoreSession: product-loop bonuses match inflected forms (asserts, fixtures, ratify)', () => {
183
+ const root = tmp('forge-score-inflect-');
184
+ try {
185
+ const { sessionDir, session } = makeSession(root, { slug: 'jobs-loop' });
186
+ fs.writeFileSync(
187
+ path.join(sessionDir, 'spine.json'),
188
+ `${JSON.stringify(validSpine([validRow()]), null, 2)}\n`,
189
+ 'utf8',
190
+ );
191
+ // Every keyword deliberately inflected — a trailing \b used to reject all of these.
192
+ fs.writeFileSync(
193
+ path.join(sessionDir, 'verify-evidence.md'),
194
+ '# Verify\n\n## Product loop\n\n1. seed fixtures\n2. run the loop\n3. asserts the ratify output differs from stored baselines\n',
195
+ 'utf8',
196
+ );
197
+ const card = scoreSession({ cwd: root, sessionDir, session });
198
+ const loop = card.checks.find((c) => c.id === 'product_loop');
199
+ const notes = loop.notes.join('\n');
200
+ assert.match(notes, /names a fixture/);
201
+ assert.match(notes, /asserts decision\/output change/);
202
+ assert.equal(loop.points, 20);
203
+ } finally {
204
+ fs.rmSync(root, { recursive: true, force: true });
205
+ }
206
+ });
207
+
208
+ test('scoreSession: green e2e run proves the loop even without the phrase', () => {
209
+ const root = tmp('forge-score-e2e-');
210
+ try {
211
+ const { sessionDir, session } = makeSession(root, { slug: 'jobs-loop' });
212
+ fs.writeFileSync(
213
+ path.join(sessionDir, 'spine.json'),
214
+ `${JSON.stringify(validSpine([validRow()]), null, 2)}\n`,
215
+ 'utf8',
216
+ );
217
+ // Evidence deliberately avoids the words "product loop" — before the fix
218
+ // this session scored 0/20 despite an executed green run.
219
+ fs.writeFileSync(
220
+ path.join(sessionDir, 'verify-evidence.md'),
221
+ '# Verify\n\n## What the evidence actually proves\n\nReverted the fix, reproduced the broken row, re-applied, ran the loop green.\n',
222
+ 'utf8',
223
+ );
224
+ const steps = [{ name: 'probe', cmd: 'node -e "console.log(1)"' }];
225
+ fs.writeFileSync(
226
+ path.join(sessionDir, 'e2e.json'),
227
+ `${JSON.stringify({ change: null, notApplicable: null, steps }, null, 2)}\n`,
228
+ 'utf8',
229
+ );
230
+ fs.writeFileSync(
231
+ path.join(sessionDir, 'e2e-results.json'),
232
+ `${JSON.stringify({ ok: true, stepsHash: e2eStepsHash(steps), steps: [{ name: 'probe', ok: true, exitCode: 0 }] }, null, 2)}\n`,
233
+ 'utf8',
234
+ );
235
+ const card = scoreSession({ cwd: root, sessionDir, session });
236
+ const loop = card.checks.find((c) => c.id === 'product_loop');
237
+ assert.ok(loop.points >= 8, `expected >= 8 loop points, got ${loop.points}`);
238
+ assert.match(loop.notes.join('\n'), /proven by execution/);
239
+ } finally {
240
+ fs.rmSync(root, { recursive: true, force: true });
241
+ }
242
+ });
243
+
244
+ test('forge cleanup harvests scorecard.json into the ledger before pruning', () => {
245
+ const root = tmp('forge-cleanup-harvest-');
246
+ try {
247
+ const { sessionDir, session } = makeSession(root, { phase: 'done' });
248
+ // A scorecard as an older (pre-ledger) forgekit would have written it.
249
+ fs.writeFileSync(
250
+ path.join(sessionDir, 'scorecard.json'),
251
+ `${JSON.stringify({
252
+ scoredAt: '2026-06-01T00:00:00.000Z',
253
+ sessionId: session.id,
254
+ slug: session.slug,
255
+ openspecChange: null,
256
+ score: 83,
257
+ maxScore: 100,
258
+ grade: 'B',
259
+ integrityOk: true,
260
+ caps: [],
261
+ checks: [{ id: 'spine', label: 'Spine matrix quality', points: 20, max: 25, notes: ['one row thin'] }],
262
+ }, null, 2)}\n`,
263
+ 'utf8',
264
+ );
265
+ execFileSync(process.execPath, [CLEANUP_SCRIPT, '--include-active'], {
266
+ cwd: root,
267
+ env: { ...process.env, FORGEKIT_FLEET_DIR: path.join(root, 'fleet') },
268
+ });
269
+ assert.equal(fs.existsSync(sessionDir), false);
270
+ const ledger = path.join(root, '.forge', 'scorecards.jsonl');
271
+ const lines = fs.readFileSync(ledger, 'utf8').split('\n').filter(Boolean);
272
+ assert.equal(lines.length, 1);
273
+ const entry = JSON.parse(lines[0]);
274
+ assert.equal(entry.sessionId, session.id);
275
+ assert.equal(entry.grade, 'B');
276
+ assert.equal(entry.deductions[0].id, 'spine');
277
+ } finally {
278
+ fs.rmSync(root, { recursive: true, force: true });
279
+ }
280
+ });
281
+
282
+ test('writeSessionScorecard writes json and md', () => {
283
+ const root = tmp('forge-score-write-');
284
+ try {
285
+ const { sessionDir, session } = makeSession(root, { slug: 'docs-only' });
286
+ fs.writeFileSync(
287
+ path.join(sessionDir, 'spine.json'),
288
+ `${JSON.stringify({ rows: [], notApplicable: 'docs-only' }, null, 2)}\n`,
289
+ 'utf8',
290
+ );
291
+ fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# ok\n', 'utf8');
292
+ const { jsonPath, mdPath, card } = writeSessionScorecard({ cwd: root, sessionDir, session });
293
+ assert.equal(fs.existsSync(jsonPath), true);
294
+ assert.equal(fs.existsSync(mdPath), true);
295
+ assert.equal(JSON.parse(fs.readFileSync(jsonPath, 'utf8')).grade, card.grade);
296
+
297
+ // Durable ledger: one line per session, re-scoring replaces (not appends).
298
+ const ledger = path.join(root, '.forge', 'scorecards.jsonl');
299
+ assert.equal(fs.existsSync(ledger), true);
300
+ writeSessionScorecard({ cwd: root, sessionDir, session });
301
+ const lines = fs.readFileSync(ledger, 'utf8').split('\n').filter(Boolean);
302
+ assert.equal(lines.length, 1);
303
+ const entry = JSON.parse(lines[0]);
304
+ assert.equal(entry.sessionId, session.id);
305
+ assert.equal(entry.grade, card.grade);
306
+ assert.ok(Array.isArray(entry.deductions));
307
+ } finally {
308
+ fs.rmSync(root, { recursive: true, force: true });
309
+ }
310
+ });
311
+
312
+ test('forge phase done writes scorecard and stamps session.scoreGrade', () => {
313
+ const root = tmp('forge-score-phase-');
314
+ try {
315
+ const { sessionDir } = makeSession(root, {
316
+ slug: 'add-health',
317
+ phase: 'verify',
318
+ tasksTotal: 1,
319
+ tasksComplete: 1,
320
+ });
321
+ fs.writeFileSync(
322
+ path.join(sessionDir, 'spine.json'),
323
+ `${JSON.stringify({ rows: [], notApplicable: 'sync HTTP only' }, null, 2)}\n`,
324
+ 'utf8',
325
+ );
326
+ fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# ok\n', 'utf8');
327
+
328
+ execFileSync(process.execPath, [PHASE_SCRIPT, 'done'], {
329
+ cwd: root,
330
+ encoding: 'utf8',
331
+ stdio: ['ignore', 'pipe', 'pipe'],
332
+ });
333
+
334
+ assert.equal(fs.existsSync(path.join(sessionDir, 'scorecard.md')), true);
335
+ assert.equal(fs.existsSync(path.join(sessionDir, 'scorecard.json')), true);
336
+ const session = JSON.parse(fs.readFileSync(path.join(sessionDir, 'session.json'), 'utf8'));
337
+ assert.equal(session.phase, 'done');
338
+ assert.ok(typeof session.score === 'number');
339
+ assert.ok(['A', 'B', 'C', 'D', 'F'].includes(session.scoreGrade));
340
+ } finally {
341
+ fs.rmSync(root, { recursive: true, force: true });
342
+ }
343
+ });
344
+
345
+ test('forge score CLI prints JSON', () => {
346
+ const root = tmp('forge-score-cli-');
347
+ try {
348
+ const { sessionDir } = makeSession(root, { slug: 'add-health' });
349
+ fs.writeFileSync(
350
+ path.join(sessionDir, 'spine.json'),
351
+ `${JSON.stringify({ rows: [], notApplicable: 'sync HTTP only' }, null, 2)}\n`,
352
+ 'utf8',
353
+ );
354
+ fs.writeFileSync(path.join(sessionDir, 'verify-evidence.md'), '# ok\n', 'utf8');
355
+ const out = execFileSync(process.execPath, [SCORE_SCRIPT], {
356
+ cwd: root,
357
+ encoding: 'utf8',
358
+ });
359
+ const card = JSON.parse(out);
360
+ assert.ok(card.grade);
361
+ assert.ok(Array.isArray(card.humanPrompts));
362
+ assert.equal(card.humanPrompts.length, 4);
363
+ } finally {
364
+ fs.rmSync(root, { recursive: true, force: true });
365
+ }
366
+ });