Loading npm security reports…

@routerhub/agent-rules@1.5.159: AI-agent control-surface risk | LPM Firewall

registry / @routerhub/agent-rules / 1.5.159

@routerhub/agent-rules@1.5.159

Shared Copilot agent rules and guidelines for RouterHub projects

AI Security Review

scanned 11d ago · by lpm-firewall-ai

LPM flags this version as an AI-agent control-surface risk. On npm install, postinstall.js changes the consumer package-manager config and runs merge.js init. That writes always-loaded Claude, Cursor, and Copilot instruction files plus Claude skills into the project, and adds a Copilot rule that tells review to ignore those files.

Static reason

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

Trigger

Automatic npm postinstall after the package is installed in a project.

Impact

Every later agent session in that project loads publisher-controlled rules and skills, review is told to skip those files, and future installs are less likely to skip the hook.

Mechanism

The hook finds the project root, appends pnpm settings that keep lifecycle scripts running and disable a frozen lockfile, then spawns merge.js init to overwrite agent instruction files and copy the skills tree.

Policy narrative

Installing the package runs postinstall.js. The script locates the consumer project and appends .npmrc entries that turn off a frozen lockfile, force dependency lifecycle scripts, and disable the side-effects cache. It then starts merge.js init, which writes AGENTS.md and CLAUDE.md, Copilot instruction files, a pull-request template, and a full copy of skills under .claude/skills. One generated Copilot file tells code review to skip those planted files. There is no prompt. Later agent sessions load this control surface automatically.

AI rationale

npm postinstall, without a consent prompt, rewrites the consumer .npmrc and plants broad Claude, Cursor, and Copilot instruction files plus Claude skills. That is unconsented install-time mutation of foreign agent control surfaces, including a rule that tells review to ignore the planted files.

Evidence

package.jsonpostinstall.jsmerge.js

Source & flagged code

6 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"

High

Suspicious Lifecycle Evidence

npm postinstall runs postinstall.js with no consent prompt.

package.jsonView on unpkg · L28

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

skills/loop-review/scripts/save-review-decision.shView file

•path = skills/loop-review/scripts/save-review-decision.sh kind = build_helper sizeBytes = 2323 magicHex = [redacted]

Medium

Ships Build Helper

Package ships non-JavaScript build or shell helper files.

skills/loop-review/scripts/save-review-decision.shView on unpkg

postinstall.jsView file

•matchType = normalized_sha256 matchedPackage = @routerhub/agent-rules@1.5.221 matchedPath = postinstall.js matchedIdentity = npm:QHJvdXRlcmh1Yi9hZ2VudC1ydWxlcw:1.5.221 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

skills/screenshot-annotate/annotate.jsView file

•matchType = normalized_sha256 matchedPackage = @routerhub/agent-rules@1.5.221 matchedPath = skills/screenshot-annotate/annotate.js matchedIdentity = npm:QHJvdXRlcmh1Yi9hZ2VudC1ydWxlcw:1.5.221 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.

skills/screenshot-annotate/annotate.jsView on unpkg