惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Last Week in AI
Last Week in AI
阮一峰的网络日志
阮一峰的网络日志
P
Proofpoint News Feed
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
MongoDB | Blog
MongoDB | Blog
云风的 BLOG
云风的 BLOG
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
J
Java Code Geeks
WordPress大学
WordPress大学
T
The Blog of Author Tim Ferriss
V
Visual Studio Blog
小众软件
小众软件
Microsoft Azure Blog
Microsoft Azure Blog
博客园_首页
IT之家
IT之家
Vercel News
Vercel News
C
Check Point Blog
Google DeepMind News
Google DeepMind News
月光博客
月光博客
D
DataBreaches.Net
酷 壳 – CoolShell
酷 壳 – CoolShell
美团技术团队
Y
Y Combinator Blog
Hugging Face - Blog
Hugging Face - Blog

Recent Commits to openclaw:main

test: merge chat side-result checks · openclaw/openclaw@ddd2c2a test: merge cron history checks · openclaw/openclaw@f7eb746 test: merge responsive navigation shell checks · openclaw/openclaw@c2e4b47 docs(changelog): add codex oauth fixes · openclaw/openclaw@628e6cd test: merge navigation routing cases · openclaw/openclaw@5d8cecb Tests: mock channel registry bundled fallback · openclaw/openclaw@2b08233 Secrets: avoid broad web search discovery for single plugin config · openclaw/openclaw@a464f59 test: merge config view browser checks · openclaw/openclaw@20cf511 fix(status): align oauth health with runtime · openclaw/openclaw@eed7116 feat: add macOS screen snapshots for monitor preview (#67954) thanks … · openclaw/openclaw@f377db1 fix: report shared auth scopes in hello-ok (#67810) thanks @BunsDev · openclaw/openclaw@0b6c39b Auto-reply: avoid eager bundled route fallback · openclaw/openclaw@3ea1bf4 Tests: narrow session binding contract setup · openclaw/openclaw@54e4e16 fix(macOS): enable undo/redo in webchat composer text input (#34962) · openclaw/openclaw@00951dc Tests: speed up channel setup promotion · openclaw/openclaw@82b529a Docs: refresh agent instructions · openclaw/openclaw@5775fe2 fix(auth): serialize OAuth refresh across agents to fix #26322 (#67876) · openclaw/openclaw@8e79080 test: allow ollama public surface boundary test · openclaw/openclaw@7d4f1a6 Docs: add test performance guardrails · openclaw/openclaw@89706d3 Tests: restore context-engine usage proof · openclaw/openclaw@e4c4f95 Tests: slim context engine runtime coverage · openclaw/openclaw@74c198f ci: retry failed custom checkouts · openclaw/openclaw@0ee5baf test: trim duplicate provider auth onboarding cases · openclaw/openclaw@1ffc02e matrix: fix sessions_spawn --thread subagent session spawning (#67643) · openclaw/openclaw@1ce2596 test: reduce auth choice fixture churn · openclaw/openclaw@857b9cd test: mock health status config boundaries · openclaw/openclaw@9d5ab4a test: mock onboard config io boundary · openclaw/openclaw@299694d test: mock legacy state plugin boundaries · openclaw/openclaw@2713089 test: mock channel install boundaries · openclaw/openclaw@b945248 test: mock doctor preview channel boundaries · openclaw/openclaw@b1a3ad4
feat(dreaming): add report-only shadow trial runner · ope...
iFiras-Max1 · 2026-05-31 · via Recent Commits to openclaw:main
1+

import fs from "node:fs/promises";

2+

import path from "node:path";

3+

import { describe, expect, it } from "vitest";

4+

import {

5+

buildDreamingShadowTrialReport,

6+

defaultDreamingShadowTrialReportPath,

7+

resolveDreamingShadowTrialRecommendation,

8+

writeDreamingShadowTrialReport,

9+

} from "./dreaming-shadow-trial.js";

10+

import { createMemoryCoreTestHarness } from "./test-helpers.js";

11+12+

const { createTempWorkspace } = createMemoryCoreTestHarness();

13+14+

const baseInput = {

15+

candidate: "The user prefers release notes with exact verification commands.",

16+

trialPrompt: "Prepare a release readiness note.",

17+

baselineOutcome: "Mentions tests passed without the exact command.",

18+

candidateOutcome: "Includes the exact verification command and remaining risk.",

19+

reason: "The candidate improves the release reply without exposing private data.",

20+

riskFlags: ["no secret exposure", "no outdated preference conflict"],

21+

evidenceRefs: ["memory/2026-05-18.md#L30-L49"],

22+

};

23+24+

describe("dreaming shadow trial runner", () => {

25+

it("maps verdicts to report-only recommendations", () => {

26+

expect(resolveDreamingShadowTrialRecommendation("helpful")).toBe("promote");

27+

expect(resolveDreamingShadowTrialRecommendation("neutral")).toBe("defer");

28+

expect(resolveDreamingShadowTrialRecommendation("harmful")).toBe("reject");

29+

});

30+31+

it("builds the stable shadow-trial report contract", () => {

32+

const report = buildDreamingShadowTrialReport({

33+

...baseInput,

34+

verdict: "helpful",

35+

nowMs: Date.parse("2026-05-18T18:00:00.000Z"),

36+

});

37+38+

expect(report.recommendation).toBe("promote");

39+

expect(report.promotionAction).toBe("report-only");

40+

expect(report.markdown).toContain("candidate: The user prefers release notes");

41+

expect(report.markdown).toContain("baseline outcome: Mentions tests passed");

42+

expect(report.markdown).toContain("candidate outcome: Includes the exact verification command");

43+

expect(report.markdown).toContain("verdict: helpful");

44+

expect(report.markdown).toContain("recommendation: promote");

45+

expect(report.markdown).toContain("risk flags:");

46+

expect(report.markdown).toContain("- no secret exposure");

47+

expect(report.markdown).toContain("evidence refs:");

48+

expect(report.markdown).toContain("promotion action: report-only");

49+

expect(report.markdown).not.toContain("promoted to MEMORY.md");

50+

});

51+52+

it("writes only the shadow-trial report and leaves MEMORY.md unchanged", async () => {

53+

const workspaceDir = await createTempWorkspace("openclaw-shadow-trial-");

54+

const memoryPath = path.join(workspaceDir, "MEMORY.md");

55+

await fs.writeFile(memoryPath, "# Memory\n\nExisting durable memory.\n", "utf-8");

56+57+

const report = await writeDreamingShadowTrialReport({

58+

...baseInput,

59+

verdict: "neutral",

60+

workspaceDir,

61+

nowMs: Date.parse("2026-05-18T18:00:00.000Z"),

62+

});

63+64+

expect(report.recommendation).toBe("defer");

65+

expect(path.dirname(report.reportPath!)).toBe(

66+

path.join(workspaceDir, "memory", "dreaming", "shadow-trials", "2026-05-18"),

67+

);

68+

expect(path.basename(report.reportPath!)).toMatch(/^[a-f0-9]{12}\.md$/);

69+

await expect(fs.readFile(memoryPath, "utf-8")).resolves.toBe(

70+

"# Memory\n\nExisting durable memory.\n",

71+

);

72+

expect(report.reportPath).toBeTruthy();

73+

await expect(fs.readFile(report.reportPath!, "utf-8")).resolves.toContain(

74+

"promotion action: report-only",

75+

);

76+

});

77+78+

it("uses the configured dreaming timezone for the default report day", async () => {

79+

const workspaceDir = await createTempWorkspace("openclaw-shadow-trial-timezone-");

80+81+

const report = await writeDreamingShadowTrialReport({

82+

...baseInput,

83+

verdict: "helpful",

84+

workspaceDir,

85+

nowMs: Date.parse("2026-05-18T21:30:00.000Z"),

86+

timezone: "Asia/Riyadh",

87+

});

88+89+

expect(path.dirname(report.reportPath!)).toBe(

90+

path.join(workspaceDir, "memory", "dreaming", "shadow-trials", "2026-05-19"),

91+

);

92+

expect(path.basename(report.reportPath!)).toMatch(/^[a-f0-9]{12}\.md$/);

93+

await expect(fs.readFile(report.reportPath!, "utf-8")).resolves.toContain(

94+

"recommendation: promote",

95+

);

96+

});

97+98+

it("keeps distinct same-day trials in separate default report files", async () => {

99+

const workspaceDir = await createTempWorkspace("openclaw-shadow-trial-collisions-");

100+

const nowMs = Date.parse("2026-05-18T18:00:00.000Z");

101+102+

const first = await writeDreamingShadowTrialReport({

103+

...baseInput,

104+

verdict: "helpful",

105+

workspaceDir,

106+

nowMs,

107+

});

108+

const second = await writeDreamingShadowTrialReport({

109+

...baseInput,

110+

candidate: "The user prefers terse release notes with exact verification commands.",

111+

verdict: "helpful",

112+

workspaceDir,

113+

nowMs,

114+

});

115+116+

expect(first.reportPath).not.toBe(second.reportPath);

117+

expect(path.dirname(first.reportPath!)).toBe(path.dirname(second.reportPath!));

118+

await expect(fs.readFile(first.reportPath!, "utf-8")).resolves.toContain(

119+

"candidate: The user prefers release notes",

120+

);

121+

await expect(fs.readFile(second.reportPath!, "utf-8")).resolves.toContain(

122+

"candidate: The user prefers terse release notes",

123+

);

124+

});

125+126+

it("keeps risky candidates reject-only without promoting durable memory", async () => {

127+

const workspaceDir = await createTempWorkspace("openclaw-shadow-trial-risk-");

128+

const reportPath = defaultDreamingShadowTrialReportPath({

129+

...baseInput,

130+

candidate: "The user always wants private tokens pasted into status reports.",

131+

candidateOutcome: "Includes a private token in the release reply.",

132+

verdict: "harmful",

133+

reason: "The candidate creates secret exposure risk.",

134+

riskFlags: ["secret exposure"],

135+

workspaceDir,

136+

nowMs: Date.parse("2026-05-19T01:00:00.000Z"),

137+

});

138+139+

const report = await writeDreamingShadowTrialReport({

140+

...baseInput,

141+

candidate: "The user always wants private tokens pasted into status reports.",

142+

candidateOutcome: "Includes a private token in the release reply.",

143+

verdict: "harmful",

144+

reason: "The candidate creates secret exposure risk.",

145+

riskFlags: ["secret exposure"],

146+

workspaceDir,

147+

reportPath,

148+

});

149+150+

expect(report.recommendation).toBe("reject");

151+

expect(report.markdown).toContain("verdict: harmful");

152+

expect(report.markdown).toContain("recommendation: reject");

153+

expect(report.markdown).toContain("promotion action: report-only");

154+

await expect(fs.readFile(path.join(workspaceDir, "MEMORY.md"), "utf-8")).rejects.toMatchObject({

155+

code: "ENOENT",

156+

});

157+

});

158+

});