Loading npm security reports…

@routerhub/agent-rules@1.5.131: AI-agent control-surface risk | LPM Firewall

registry / @routerhub/agent-rules / 1.5.131

@routerhub/agent-rules@1.5.131

Shared Copilot agent rules and guidelines for RouterHub projects

AI Security Review

scanned 2d ago · by lpm-firewall-ai

LPM flags this version as an AI-agent control-surface risk. On npm postinstall, the package alters the consuming project's npm configuration and generates AI-agent instruction files. It also installs package-supplied skills into the project's Claude control directory.

Static reason

One or more suspicious static signals were detected.; source matched previously finalized malicious package; routed for review; source fingerprint signature matched known malicious package; routed for review

Trigger

npm installation runs postinstall

Impact

Package-controlled instructions and skills become active in the consumer project's AI-agent surfaces without an explicit user command.

Mechanism

automatic foreign project AI-agent configuration and skill deployment

Policy narrative

Installing the dependency runs postinstall, finds the consumer project, changes its .npmrc to ensure lifecycle execution, and launches initialization. Initialization writes AGENTS.md, CLAUDE.md, Copilot instruction files, and package-provided skills under .claude/skills. These are behavior-bearing AI-agent control surfaces in a foreign project and are mutated without an explicit user setup command.

Rationale

This is a concrete, unconsented postinstall mutation of broad foreign AI-agent control surfaces, which meets the blocking policy. No network exfiltration is needed for the control-hijack behavior.

Evidence

package.jsonpostinstall.jsmerge.js.npmrcAGENTS.mdCLAUDE.mdAGENTS.private.md.github/copilot-instructions.md.github/instructions/*.instructions.md

Source & flagged code

7 flagged · loading source

package.jsonView file

•scripts.postinstall = node postinstall.js

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg

•scripts.postinstall = node postinstall.js

Medium

Ambiguous Install Lifecycle Script

Install-time lifecycle script is not statically allowlisted and needs review.

package.jsonView on unpkg

merge.jsView file

•Manifest-trigger-reachable source links an external AI-agent control path to a behavior-bearing write operation. sole.log(`✅ 规则已合并到 ${outputpath}`); } // 同步 skills syncskills(config); return; } // 2. 读取并解析私有规则的域标记 const privatecontent = readfileifexists(privaterulespath); const privatedomains = parseprivatedomains(privatecontent); // 3. 拼接全局规则内容 const globalcontent = globalrules .map((rule) => rule.content.trim()) .join("\n\n"); // 4. 生成 .github/copilot-instructions.md（全局规则 + 私有全局部分） let copilotcontent = "# copilot agent rules\n\n" + globalcontent; if (privatedomains.global) { copilotcontent += "\n\n---\n\n# 项目私有规则\n\n" + privatedomains.global; } ensureparentdir(copilotoutput); fs.writefilesync(copilotoutput, copilotcontent, "utf-8");

High

Trigger Reachable External Ai Agent Control Surface Mutation

Manifest-trigger-reachable source writes behavior-bearing configuration into a user or project AI-agent control surface.

merge.jsView on unpkg

postinstall.jsView file

•matchType = normalized_sha256 matchedPackage = @routerhub/agent-rules@1.5.123 matchedPath = postinstall.js matchedIdentity = npm:QHJvdXRlcmh1Yi9hZ2VudC1ydWxlcw:1.5.123 similarity = 1.000 summary = normalized source hash matched finalized malicious source

High

Known Malware Source Similarity

Source file is highly similar to a previously finalized malicious package; route for source-aware review.

postinstall.jsView on unpkg

•matchType = malicious_source_fingerprint_signature signature = 237cac60893b7f35 signatureType = suspicious_hashes sourceLabel = final_verdict:malicious matchedPackage = @routerhub/agent-rules@1.5.123 matchedPath = postinstall.js matchedIdentity = npm:QHJvdXRlcmh1Yi9hZ2VudC1ydWxlcw:1.5.123 similarity = 1.000 shingleOverlap = 2 summary = package final verdict is malicious

High

Known Malware Source Fingerprint Signature

Source fingerprint signature matches a known malicious package signature; route for source-aware review.

postinstall.jsView on unpkg

70const npmrcPath = path.join(projectRoot, ".npmrc"); let npmrcContent = ""; if (fs.existsSync(npmrcPath)) { npmrcContent = fs.readFileSync(npmrcPath, "utf-8"); } const npmrcEntries = [];

High

Suspicious Lifecycle Evidence

postinstall locates the consuming project and appends pnpm execution settings to its .npmrc.

postinstall.jsView on unpkg · L70

Source & flagged code

7 flagged

package.jsonView file

Lines 8-46json

8 },

9 "files": [

10 "AGENTS.base.md",

11 "rules/",

12 "skills/",

13 "merge.js",

14 "AGENTS.private.example.md",

15 "PULL_REQUEST_TEMPLATE.md",

16 "README.zh-CN.md",

17 "postinstall.js",

18 "package.json",

19 "CHANGELOG.md"

20 ],

21 "scripts": {

22 "build": "node --check merge.js && node --check postinstall.js",

23 "test:postinstall-init": "node tests/postinstall-init.test.js",

24 "merge": "node merge.js",

25 "sync": "node merge.js sync",

26 "init": "node merge.js init",

27 "watch": "node merge.js watch",

28 "postinstall": "node postinstall.js"

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg · L28

Medium

Ambiguous Install Lifecycle Script

Install-time lifecycle script is not statically allowlisted and needs review.

package.jsonView on unpkg · L28

29 },

30 "repository": {

31 "type": "git",

32 "url": "git+https://github.com/PomexAITeam/agent-rules.git"

33 },

34 "keywords": [

35 "copilot",

36 "agent",

37 "rules",

38 "guidelines"

39 ],

40 "author": "RouterHub Team",

41 "license": "MIT",

42 "engines": {

43 "node": ">=16.0.0"

44 }

45}

46

merge.jsView file

•Manifest-trigger-reachable source links an external AI-agent control path to a behavior-bearing write operation. sole.log(`✅ 规则已合并到 ${outputpath}`); } // 同步 skills syncskills(config); return; } // 2. 读取并解析私有规则的域标记 const privatecontent = readfileifexists(privaterulespath); const privatedomains = parseprivatedomains(privatecontent); // 3. 拼接全局规则内容 const globalcontent = globalrules .map((rule) => rule.content.trim()) .join("\n\n"); // 4. 生成 .github/copilot-instructions.md（全局规则 + 私有全局部分） let copilotcontent = "# copilot agent rules\n\n" + globalcontent; if (privatedomains.global) { copilotcontent += "\n\n---\n\n# 项目私有规则\n\n" + privatedomains.global; } ensureparentdir(copilotoutput); fs.writefilesync(copilotoutput, copilotcontent, "utf-8");

High

Trigger Reachable External Ai Agent Control Surface Mutation

Manifest-trigger-reachable source writes behavior-bearing configuration into a user or project AI-agent control surface.

merge.jsView on unpkg

postinstall.jsView file

Lines 50-115javascript

50}

51

52// 获取下游项目根目录

53const projectRoot = findProjectRoot();

54

55// 安全检查：如果项目根目录仍在 node_modules 内，跳过初始化

56if (projectRoot.includes("node_modules")) {

57 console.