registry / Install hook abuse / yu-method

yu-method

YU-METHOD — personal agile AI development skills built on BMAD-METHOD with customizable hooks

This report shows the evidence for yu-method@1.1.4. 3 of 3 known versions are blocked or flagged. See all versions.

AI Security Review

scanned 2mo ago · by lpm-firewall-ai

The automatic lifecycle hook makes broad, persistent changes to foreign AI-agent control surfaces and installs an additional global package. This meets the install-hook abuse boundary regardless of the absence of direct credential theft.

Static reason

High-risk behavior combination matched malicious policy.; source matched previously finalized malicious package; routed for review; source fingerprint signature matched known malicious package; routed for review

Trigger

npm installation of yu-method

Impact

The package changes how Claude, Codex, and other agent environments load skills and respond to future prompts without an explicit setup command.

Mechanism

Postinstall mutation of global agent skills, instructions, and MCP configuration

Attack narrative

On npm installation, package.json invokes scripts/install.js. That hook copies its bundled skills into global directories for multiple AI-agent products, installs a separate global package, registers it as a Claude MCP server, and appends rules to Claude and Codex instruction files. The inserted rules direct agents to invoke package skills automatically before handling certain future user messages. These are broad, persistent AI-agent control-surface changes performed without an explicit user setup action.

AI rationale

Evidence

package.jsonscripts/install.js~/.claude/skills/yu-*~/.agents/skills/yu-*~/.codex/skills/yu-*~/.claude.json~/.claude/CLAUDE.md~/.codex/AGENTS.md~/AGENTS.md

Decision evidence

public snapshot

AI called this Malicious at 99.0% confidence as Malware with low false-positive risk.

Evidence for block

The postinstall hook automatically runs scripts/install.js.
Installation copies bundled skills into existing Claude, Agents, and Codex global skill folders.
Installation silently installs a global CodeGraph package and registers it as a Claude MCP server.
Installation appends automatic skill-invocation rules to Claude and Codex instruction files, including the home-level AGENTS.md.

Evidence against

The inspected automatic installer contains no credential collection or direct data-exfiltration code.
The gbrain setup script is exposed through an explicit command rather than the postinstall hook.
No runtime dependency on yu-method itself is declared in package.json.

Behavioral surface

Source

ChildProcessEnvironmentVarsFilesystem

Supply chain

HighEntropyStringsUrlStrings

ManifestNo manifest risk signals triggered.

scanned 8 file(s), 53.6 KB of source, external domains: ollama.com

Source & flagged code

10 flagged

package.jsonView file

Lines 21-51json

21 "bugs": {

22 "url": "https://github.com/BestHenryWu/yu-method/issues"

23 },

24 "private": false,

25 "files": [

26 ".claude-plugin/",

27 "scripts/",

28 "skills/",

29 "team-brain/",

30 "docs/",

31 "WORKFLOW.md",

32 "CHANGELOG.md",

33 "module-help.csv",

34 "README.md"

35 ],

36 "bin": {

37 "yu-method": "scripts/cli.js",

38 "yu-brain-setup": "scripts/brain-setup.js"

39 },

40 "scripts": {

41 "postinstall": "node scripts/install.js",

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg · L41

Medium

Ambiguous Install Lifecycle Script

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

package.jsonView on unpkg · L41

42 "sync": "node scripts/sync.js",

43 "validate": "node scripts/validate-toml.js",

44 "check:bmad": "node scripts/check-bmad-compat.js",

45 "prepublishOnly": "node scripts/validate-toml.js"

46 },

47 "engines": {

48 "node": ">=20.0.0"

49 }

50}

scripts/brain-setup.jsView file

Lines 1-26javascript

1#!/usr/bin/env node

2'use strict';

4const fs = require('fs');

5const path = require('path');

scripts/install.jsView file

Lines 1-28javascript

1#!/usr/bin/env node

2'use strict';

4const fs = require('fs');

5const path = require('path');

scripts/session-end-hook.shView file

•path = scripts/session-end-hook.sh kind = build_helper sizeBytes = 1150 magicHex = [redacted]

Medium

Ships Build Helper

Package ships non-JavaScript build or shell helper files.

scripts/session-end-hook.shView on unpkg

scripts/brain-setup.js#virtual:normalized:round1View file

•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 2

High

Semantic Analysis Limited

A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.

scripts/brain-setup.js#virtual:normalized:round1View on unpkg

Source & flagged code

10 flagged · loading source

package.jsonView file

•scripts.postinstall = node scripts/install.js

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg

•scripts.postinstall = node scripts/install.js

Medium

Ambiguous Install Lifecycle Script

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

package.jsonView on unpkg

scripts/brain-setup.jsView file

6const os = require('os'); L7: const { spawnSync } = require('child_process'); L8: const readline = require('readline');

High

Child Process

Package source references child process execution.

scripts/brain-setup.jsView on unpkg · L6

Findings

1 Critical7 High4 Medium5 Low

CriticalAi Agent Control Hijackscripts/install.js

HighInstall Time Lifecycle Scriptspackage.json

HighChild Processscripts/brain-setup.js

HighTrigger Reachable External Ai Agent Control Surface Mutationscripts/install.js

HighRuntime Package Installscripts/install.js

HighSemantic Analysis Limitedscripts/brain-setup.js#virtual:normalized:round1

HighKnown Malware Source Similarityscripts/brain-setup.js

HighKnown Malware Source Fingerprint Signaturescripts/brain-setup.js

MediumAmbiguous Install Lifecycle Scriptpackage.json

MediumEnvironment Vars

MediumShips Build Helperscripts/session-end-hook.sh

Affected versions and remediation

This report applies to yu-method@1.1.4.

The versions list below shows the verdict for each recorded version.

Avoid installing yu-method@1.1.4. Remove it from direct dependencies and check your lockfile for transitive copies.
Choose an independently verified alternative or release. This report does not establish that other versions are safe.
If this version ran, investigate the affected machine and build environment. Rotate credentials it could access and rebuild from a trusted environment.

Evidence last updated: 2026-09-07 09:04:39.414Z (UTC).

Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.

Versions

3 known versions

1.1.4latestshownMaliciousInstall Hook AbuseBlockPublishedSep 6, 2026, 5:16 AM 1.1.1MaliciousSoft block: AI-agent control surfaceBlockPublishedSep 3, 2026, 1:13 PM 1.0.0MaliciousSoft block: AI-agent control surfaceBlockPublishedSep 1, 2026, 12:26 PM

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Affected versions and remediation

Versions

Related packages

Other install hook abuse packages

Behavioral surface

Related packages

Other install hook abuse packages

Source & flagged code