rungate 0.2.0

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 (196) hide show
  1. package/.claude/agents/aditi.md +47 -0
  2. package/.claude/agents/marcus.md +50 -0
  3. package/.claude/agents/quinn.md +58 -0
  4. package/.claude/agents/rook.md +47 -0
  5. package/.claude/agents/serena.md +46 -0
  6. package/.claude/rungate.json +11 -0
  7. package/.fallowrc.json +28 -0
  8. package/.github/copilot-instructions.md +20 -0
  9. package/.rungate/conformity-findings.json +648 -0
  10. package/AGENTS.md +106 -0
  11. package/CODE-MAP.md +50 -0
  12. package/HARNESS.md +88 -0
  13. package/bun.lock +26 -0
  14. package/bunfig.toml +2 -0
  15. package/config/brief-policies.json +28 -0
  16. package/config/ceremony-profiles.json +212 -0
  17. package/config/consumer-cascade-map.json +24 -0
  18. package/config/doc-cascade-map.json +24 -0
  19. package/config/eval.json +84 -0
  20. package/config/profiles/account-plan.yaml +36 -0
  21. package/config/profiles/campaign.yaml +94 -0
  22. package/config/prove-evidence.schema.json +65 -0
  23. package/config/spec-policies.json +36 -0
  24. package/config/workflow-schema.json +281 -0
  25. package/gates/SCHEMA-GUIDE.md +121 -0
  26. package/gates/ac-quality.test.ts +106 -0
  27. package/gates/adversarial.test.ts +274 -0
  28. package/gates/brief-assembler.test.ts +429 -0
  29. package/gates/brief-assembler.ts +334 -0
  30. package/gates/bunfig.toml +2 -0
  31. package/gates/chain.test.ts +91 -0
  32. package/gates/e2e-smoke.test.ts +234 -0
  33. package/gates/error-classifier.test.ts +249 -0
  34. package/gates/error-classifier.ts +89 -0
  35. package/gates/orchestrator.test.ts +950 -0
  36. package/gates/orchestrator.ts +569 -0
  37. package/gates/preload.ts +77 -0
  38. package/gates/prompts/ac-adversary.md +20 -0
  39. package/gates/prompts/evidence-validator.md +22 -0
  40. package/gates/prompts/prove-reproducer.md +53 -0
  41. package/gates/prove.test.ts +164 -0
  42. package/gates/run-gate.ts +936 -0
  43. package/gates/schema-parity.test.ts +197 -0
  44. package/gates/schema.ts +362 -0
  45. package/gates/self-heal.test.ts +511 -0
  46. package/gates/self-heal.ts +121 -0
  47. package/gates/ship-orchestrator.test.ts +379 -0
  48. package/gates/ship-orchestrator.ts +363 -0
  49. package/gates/test-fixtures/broken-no-environments.json +60 -0
  50. package/gates/test-fixtures/valid-standard.json +54 -0
  51. package/gates/test-utils.ts +43 -0
  52. package/gates/witness.ts +220 -0
  53. package/gates/workflow.test.ts +1053 -0
  54. package/hooks/AgentBriefGuard.hook.ts +586 -0
  55. package/hooks/AgentVerdictCapture.hook.ts +134 -0
  56. package/hooks/AutoVerifyGate.hook.ts +76 -0
  57. package/hooks/GateEnforcement.hook.ts +273 -0
  58. package/hooks/IssueCloseGuard.hook.ts +116 -0
  59. package/hooks/MergeGuard.hook.ts +55 -0
  60. package/hooks/StaleTTLCleanup.hook.ts +133 -0
  61. package/hooks/VerifyPhaseLock.hook.ts +81 -0
  62. package/hooks/WorkflowStateGuard.hook.ts +37 -0
  63. package/hooks/lib/agentDetection.ts +70 -0
  64. package/hooks/lib/findWorkflow.ts +101 -0
  65. package/hooks/lib/parseStdin.ts +37 -0
  66. package/hooks/lib/paths.ts +111 -0
  67. package/hooks/lib/utils.ts +81 -0
  68. package/lib/conformity.ts +959 -0
  69. package/lib/paths.ts +23 -0
  70. package/lib/rungate-schema.ts +60 -0
  71. package/lib/signal-phrases.ts +7 -0
  72. package/package.json +20 -0
  73. package/prompts/ac-adversary.md +20 -0
  74. package/prompts/ac-format.md +34 -0
  75. package/prompts/aditi.md +29 -0
  76. package/prompts/blast-radius.md +29 -0
  77. package/prompts/coding-principles.md +34 -0
  78. package/prompts/container-rebuild.md +29 -0
  79. package/prompts/container-verify.md +28 -0
  80. package/prompts/discovery.md +27 -0
  81. package/prompts/environment.md +28 -0
  82. package/prompts/escalation-decision-tree.md +38 -0
  83. package/prompts/evidence-hierarchy.md +31 -0
  84. package/prompts/evidence-validator.md +22 -0
  85. package/prompts/marcus.md +29 -0
  86. package/prompts/prevention.md +26 -0
  87. package/prompts/prove-reproducer.md +43 -0
  88. package/prompts/quinn-decision-tree.md +31 -0
  89. package/prompts/quinn-ui-brief.md +106 -0
  90. package/prompts/quinn.md +29 -0
  91. package/prompts/rca.md +28 -0
  92. package/prompts/read-before-write.md +30 -0
  93. package/prompts/regression.md +26 -0
  94. package/prompts/rook.md +29 -0
  95. package/prompts/serena.md +28 -0
  96. package/prompts/testing-strategy.md +34 -0
  97. package/reference/migration-manifest.json +554 -0
  98. package/reference/prompts/agent-aditi.md +38 -0
  99. package/reference/prompts/agent-marcus.md +38 -0
  100. package/reference/prompts/agent-quinn.md +38 -0
  101. package/reference/prompts/agent-rook.md +37 -0
  102. package/reference/prompts/agent-serena.md +37 -0
  103. package/reference/prompts/council-skill.md +170 -0
  104. package/reference/prompts/goal-skill.md +235 -0
  105. package/reference/prompts/prove-skill.md +283 -0
  106. package/reference/prompts/quinn-standard.md +246 -0
  107. package/reference/prompts/ship-convergence.md +65 -0
  108. package/reference/prompts/ship-escalation.md +43 -0
  109. package/reference/prompts/ship-iteration.md +382 -0
  110. package/reference/prompts/ship-pending-patches.md +246 -0
  111. package/reference/prompts/ship-reference.md +250 -0
  112. package/reference/prompts/ship-skill.md +347 -0
  113. package/reference/prompts/ship-templates.md +215 -0
  114. package/reference/scripts/batch-status.sh +68 -0
  115. package/reference/scripts/chain-compat-check.sh +67 -0
  116. package/reference/scripts/context-audit.sh +314 -0
  117. package/reference/scripts/dead-export-check.sh +61 -0
  118. package/reference/scripts/generate-qc-tables.sh +92 -0
  119. package/reference/scripts/harness-init.sh +111 -0
  120. package/reference/scripts/pai-054-baseline-check.sh +40 -0
  121. package/reference/scripts/prove-smoke-test.sh +112 -0
  122. package/reference/scripts/setup-keychain-salt.sh +12 -0
  123. package/reference/scripts/skill-optimizer-friday.sh +88 -0
  124. package/reference/scripts/skill-qc-test.sh +351 -0
  125. package/reference/scripts/skill-synthesize-tonight.sh +66 -0
  126. package/reference/specs/HARNESS-EXTRACTION-SPEC.md +263 -0
  127. package/reference/specs/harness-v3-bash-deletion-plan.md +65 -0
  128. package/reference/specs/harness-v3-migration-plan.md +188 -0
  129. package/scripts/decision-reconcile.sh +74 -0
  130. package/scripts/decision-reconcile.ts +152 -0
  131. package/scripts/extract-constraints.ts +260 -0
  132. package/scripts/generate-code-map.ts +339 -0
  133. package/scripts/git-hooks/pre-commit +76 -0
  134. package/scripts/git-hooks/pre-push +271 -0
  135. package/scripts/harness-skill-check.sh +52 -0
  136. package/scripts/lib/gap-emit.sh +65 -0
  137. package/scripts/lib/jq-update.sh +30 -0
  138. package/scripts/lib/resolve-slug.sh +30 -0
  139. package/scripts/lib/verify-spec-compliance.sh +102 -0
  140. package/scripts/prove-backfill.sh +87 -0
  141. package/scripts/scaffold-project.ts +1488 -0
  142. package/scripts/scaffold-rungate-config.ts +91 -0
  143. package/scripts/self-containment-check.sh +109 -0
  144. package/scripts/skill-qc-validator.sh +402 -0
  145. package/scripts/slug-resolver.sh +14 -0
  146. package/scripts/sync-spec-tests.ts +236 -0
  147. package/specs/BOOTSTRAP-DATA-FLOW-SPEC.md +2493 -0
  148. package/specs/BOOTSTRAP-TEST-PLAN.md +571 -0
  149. package/specs/HARNESS-GATES.md +256 -0
  150. package/specs/HARNESS-SKILL-CHAIN.md +286 -0
  151. package/specs/HARNESS-SKILL-CONTRACT.md +93 -0
  152. package/specs/HARNESS-STANDARD.md +488 -0
  153. package/specs/SPEC-TEMPLATE.md +53 -0
  154. package/specs/bootstrap-data-flow.png +0 -0
  155. package/specs/harness-automation-matrix.md +314 -0
  156. package/specs/harness-automation-matrix.rules.json +64 -0
  157. package/test/anti.test.ts +80 -0
  158. package/test/contract-negative.test.ts +86 -0
  159. package/test/contract.test.ts +158 -0
  160. package/test/external-deps.test.ts +48 -0
  161. package/test/fixtures/golden-project/.env.example +3 -0
  162. package/test/fixtures/golden-project/Makefile +7 -0
  163. package/test/fixtures/golden-project/docs/adr/ADR-001-framework.md +23 -0
  164. package/test/fixtures/golden-project/docs/setup-guide.md +10 -0
  165. package/test/fixtures/golden-project/package.json +10 -0
  166. package/test/fixtures/golden-project/specs/api-spec.md +20 -0
  167. package/test/fixtures/golden-project/src/auth.ts +8 -0
  168. package/test/fixtures/golden-project/src/db.ts +11 -0
  169. package/test/fixtures/golden-project/src/index.ts +19 -0
  170. package/test/fixtures/golden-project/src/utils/format.ts +3 -0
  171. package/test/fixtures/golden-project/test/auth.test.ts +15 -0
  172. package/test/fixtures/golden-project/tsconfig.json +12 -0
  173. package/test/meta-sc-coverage.test.ts +159 -0
  174. package/test/phase-0.test.ts +846 -0
  175. package/test/phase-1.test.ts +617 -0
  176. package/test/phase-2.test.ts +620 -0
  177. package/test/phase-3.test.ts +99 -0
  178. package/test/scaffold-conformity.test.ts +8 -0
  179. package/test/schema-canary.test.ts +51 -0
  180. package/test/spec-compliance-auto.test.ts +64 -0
  181. package/test/spec-compliance.test.ts +295 -0
  182. package/test/spec-discovery.test.ts +68 -0
  183. package/test/spec-drift.test.ts +5 -0
  184. package/test/structure.test.ts +165 -0
  185. package/test/unit/agent-brief-guard-ship.test.ts +293 -0
  186. package/test/unit/extract-constraints.test.ts +145 -0
  187. package/test/unit/issue-close-guard.test.ts +204 -0
  188. package/test/unit/scaffold-project.test.ts +417 -0
  189. package/test/unit/skill-cooccurrence.test.ts +177 -0
  190. package/test/unit/skill-sequence-logger.test.ts +244 -0
  191. package/test/unit/workflow-state-guard.test.ts +95 -0
  192. package/workflows/batch-ship.js +295 -0
  193. package/workflows/council.js +481 -0
  194. package/workflows/prove.js +560 -0
  195. package/workflows/ship.js +789 -0
  196. package/workflows/verify.js +236 -0
@@ -0,0 +1,53 @@
1
+ ---
2
+ doc-type: reference
3
+ status: active
4
+ owner: jason
5
+ updated: 2026-09-19
6
+ ---
7
+
8
+ You are a prove reproducer. Your objective: reproduce the bug described
9
+ in the issue, verify the fix works, and capture evidence.
10
+
11
+ **Tool limitation:** You have Bash and Read tools only — no browser,
12
+ no Playwright, no screenshot capability. For UI-related ACs (screenshots,
13
+ visual elements, page navigation, OUTCOME-type criteria), output a SKIP
14
+ verdict with reason "B3 limitation: no browser tools — requires Quinn".
15
+ Focus your verification on CODE ACs that can be tested via command line
16
+ (grep, bun test, bun -e, curl, file reads).
17
+
18
+ 1. Read the issue body (provided). Identify the user's complaint and
19
+ derive a test plan from the issue description.
20
+ 2. Compare your test plan against the ACs provided. If the issue
21
+ describes a bug the ACs don't test, flag the gap.
22
+ 3. For API/data issues: call the endpoints described in the issue via
23
+ curl, capture the response bodies as evidence.
24
+ 4. For CODE issues: run evidence commands (grep, bun test, file reads)
25
+ to verify the fix is present and correct.
26
+ 5. For UI issues: output SKIP for those criteria — you cannot navigate
27
+ pages or take screenshots. Note which ACs need Quinn verification.
28
+ 6. Compare the before-state (provided) against the current after-state.
29
+ The before state shows what was broken. The after state must show
30
+ the fix working.
31
+ 6b. For S-tier evidence (bug fixes): apply negative control — revert the
32
+ fix temporarily, confirm the bug returns, then re-apply. This proves
33
+ the fix is the actual cause of the improvement, not a coincidence.
34
+ 7. For each criterion, capture real evidence — API responses, console
35
+ output, grep results. Never claim UI verification without browser tools.
36
+
37
+ Output JSON:
38
+ {
39
+ "verdict": "PROVEN|UNPROVEN|INCONCLUSIVE",
40
+ "criteriaResults": [
41
+ {"scId": "SC-1", "verdict": "PASS|FAIL", "evidence": "description of what was observed"}
42
+ ],
43
+ "reproduced": true|false,
44
+ "evidence": [
45
+ {"type": "screenshot|api-response|console", "description": "what this shows"}
46
+ ],
47
+ "gaps": []
48
+ }
49
+
50
+ Verdict rules:
51
+ - PROVEN: bug was reproduced in before-state AND fix verified in after-state
52
+ - UNPROVEN: fix does not resolve the issue (evidence shows problem persists)
53
+ - INCONCLUSIVE: could not reproduce the bug or connect to dev server
@@ -0,0 +1,164 @@
1
+ import { test, expect, describe } from "bun:test";
2
+ import { readFileSync, existsSync, writeFileSync } from "fs";
3
+ import { execSync } from "child_process";
4
+
5
+ const TEST_DIR = process.env.TEST_WORK_DIR || "";
6
+ const PROJECT_ROOT = process.env.PROJECT_ROOT || process.cwd();
7
+ const GATE = process.env.GATE || "";
8
+
9
+ function tryRead(filename: string) {
10
+ const path = `${TEST_DIR}/${filename}`;
11
+ if (!existsSync(path)) return null;
12
+ return JSON.parse(readFileSync(path, "utf-8"));
13
+ }
14
+
15
+ function runEvidence(command: string): { exitCode: number; output: string } {
16
+ try {
17
+ const output = execSync(command, { cwd: PROJECT_ROOT, timeout: 30000, encoding: "utf-8" });
18
+ return { exitCode: 0, output: output.trim() };
19
+ } catch (e: any) {
20
+ return { exitCode: e.status ?? 1, output: (e.stdout ?? "").trim() };
21
+ }
22
+ }
23
+
24
+ function compareThreshold(actual: string, op: string, expected: string): boolean {
25
+ if (op === "contains") return actual.includes(String(expected));
26
+ if (op === "!contains") return !actual.includes(String(expected));
27
+ if (op === "exists") return actual.length > 0;
28
+ const num = parseFloat(actual);
29
+ const exp = parseFloat(expected);
30
+ if (isNaN(num) || isNaN(exp)) return actual === expected;
31
+ switch (op) {
32
+ case ">=": return num >= exp;
33
+ case "<=": return num <= exp;
34
+ case "==": return num === exp;
35
+ case ">": return num > exp;
36
+ case "<": return num < exp;
37
+ case "!=": return num !== exp;
38
+ default: return actual === expected;
39
+ }
40
+ }
41
+
42
+ describe("prove: run evidence commands from workflow-state", () => {
43
+ if (GATE && GATE !== "prove") {
44
+ test("skipped — prove tests only run for prove/ship gates", () => {});
45
+ return;
46
+ }
47
+ const wf = tryRead("workflow-state.json");
48
+ if (!wf) {
49
+ test("skipped — no active workflow (TEST_WORK_DIR not set or workflow-state.json missing)", () => {});
50
+ return;
51
+ }
52
+
53
+ const commitSha = execSync("git rev-parse --short HEAD", { cwd: PROJECT_ROOT, encoding: "utf-8" }).trim();
54
+ const ceremonyTier = wf.sizing?.ceremonyTier || "STANDARD";
55
+ const proveEnvironment = ceremonyTier === "LIGHT" ? "local" : "prod";
56
+ const criteriaResults: any[] = [];
57
+
58
+ // B3 prove reproducer has Bash+Read only — no browser/screenshot capability.
59
+ // OUTCOME-type ACs and UI-related evidence commands get SKIP, not FAIL (#486).
60
+ const B3_UI_PATTERN = /screenshot|browser|playwright|visual|navigate|page\./i;
61
+ const isB3UiAc = (ac: any): boolean => {
62
+ if (ac.type === "OUTCOME") return true;
63
+ const cmd = ac.evidenceMethod?.command || "";
64
+ return B3_UI_PATTERN.test(cmd);
65
+ };
66
+
67
+ for (const ac of wf.acs) {
68
+ test(`${ac.id}: ${ac.statement.slice(0, 80)}`, () => {
69
+ // B3 exempt: OUTCOME/UI ACs cannot be verified without browser tools
70
+ if (isB3UiAc(ac)) {
71
+ criteriaResults.push({
72
+ scId: ac.id,
73
+ verdict: "SKIP",
74
+ evidence: "B3 cannot verify UI — requires Quinn (#486)",
75
+ });
76
+ return;
77
+ }
78
+
79
+ const cmd = ac.evidenceMethod?.command;
80
+ if (!cmd) {
81
+ criteriaResults.push({ scId: ac.id, verdict: "SKIP", evidence: "no evidence command" });
82
+ return;
83
+ }
84
+
85
+ const result = runEvidence(cmd);
86
+ const lastLine = result.output.split("\n").pop() || "";
87
+ const isTestRunner = /^bun test\b/.test(cmd);
88
+
89
+ let verdict: string;
90
+ let evidence: string;
91
+
92
+ if (isTestRunner) {
93
+ verdict = result.exitCode === 0 ? "PASS" : "FAIL";
94
+ evidence = `${cmd} → exit ${result.exitCode} (test runner: exit code is verdict)`;
95
+ } else if (ac.threshold && result.exitCode === 0) {
96
+ const passed = compareThreshold(lastLine, ac.threshold.op, ac.threshold.value);
97
+ verdict = passed ? "PASS" : "FAIL";
98
+ evidence = `${cmd} → "${lastLine}" ${ac.threshold.op} ${ac.threshold.value} = ${passed}`;
99
+ } else if (result.exitCode === 0) {
100
+ verdict = "PASS";
101
+ evidence = `${cmd} → exit 0`;
102
+ } else {
103
+ verdict = "FAIL";
104
+ evidence = `${cmd} → exit ${result.exitCode}, output: "${lastLine}"`;
105
+ }
106
+
107
+ criteriaResults.push({ scId: ac.id, verdict, evidence });
108
+
109
+ if (verdict === "FAIL") {
110
+ expect(`${ac.id}: ${evidence}`).toBe("PASS");
111
+ }
112
+ });
113
+ }
114
+
115
+ test("write prove-evidence.json", () => {
116
+ const fails = criteriaResults.filter(r => r.verdict === "FAIL");
117
+ const verdict = criteriaResults.length === 0 ? "INCONCLUSIVE"
118
+ : fails.length > 0 ? "UNPROVEN" : "PROVEN";
119
+
120
+ const evidence = {
121
+ issueNumber: wf.issue,
122
+ verdict,
123
+ commitSHA: commitSha,
124
+ capturedAt: new Date().toISOString(),
125
+ beforeEvidence: wf.beforeState ?? {
126
+ type: "none", path: null,
127
+ capturedAt: new Date().toISOString(),
128
+ description: "no before-state captured",
129
+ },
130
+ afterEvidence: {
131
+ type: "text", path: null,
132
+ capturedAt: new Date().toISOString(),
133
+ environment: proveEnvironment,
134
+ description: criteriaResults.map(r => `${r.scId}: ${r.verdict}`).join(", "),
135
+ },
136
+ comparisonSummary: `${criteriaResults.filter(r => r.verdict === "PASS").length}/${criteriaResults.length} criteria passed`,
137
+ reproduced: true,
138
+ gaps: fails.map(f => ({ ac: f.scId, status: "UNPROVEN", detail: f.evidence })),
139
+ criteriaResults,
140
+ };
141
+
142
+ writeFileSync(`${TEST_DIR}/prove-evidence.json`, JSON.stringify(evidence, null, 2));
143
+ expect(verdict).toBe("PROVEN");
144
+ });
145
+
146
+ test("prod environment enforcement: STANDARD+ requires prod evidence (#294, ported to Bun in #484)", () => {
147
+ if (ceremonyTier === "LIGHT") {
148
+ // LIGHT tier is exempt from prod enforcement — API check only
149
+ return;
150
+ }
151
+
152
+ // For STANDARD and THOROUGH tiers, afterEvidence.environment must be "prod"
153
+ const evidencePath = `${TEST_DIR}/prove-evidence.json`;
154
+ if (!existsSync(evidencePath)) {
155
+ expect("prove-evidence.json").toBe("must exist before prod enforcement check");
156
+ return;
157
+ }
158
+
159
+ const proveEvidence = JSON.parse(readFileSync(evidencePath, "utf-8"));
160
+ const env = proveEvidence.afterEvidence?.environment;
161
+
162
+ expect(env).toBe("prod");
163
+ });
164
+ });