registry / AI-agent control-surface risk / @doist/twist-cli

@doist/twist-cli

TypeScript CLI for Twist

This report shows the evidence for @doist/twist-cli@3.0.8. 2 of 14 known versions are blocked or flagged. The newest version, 3.0.15, is flagged. See all versions.

AI Security Review

scanned 1mo ago · by lpm-firewall-ai

The postinstall lifecycle performs unconsented mutation of foreign and broad AI-agent control surfaces. This meets the install-hook abuse blocking boundary despite skipping paths where no prior skill exists.

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 postinstall during package installation

Impact

Changes instructions available to installed AI agents, including broad .agents support.

Mechanism

automatic overwrite of detected agent skill files

Policy narrative

When npm installs this package, its postinstall hook loads the compiled installer. That installer scans supported agent directories and overwrites every already-installed Twist skill, including Codex, Claude Code, Cursor, and .agents. The mutation is automatic and occurs outside an explicit skill-management command.

AI rationale

Evidence

package.jsonscripts/postinstall.jsdist/postinstall.jsdist/lib/skills/update-installed.jsdist/lib/skills/index.jsdist/lib/skills/create-installer.js.claude/skills/twist-cli/SKILL.md.codex/skills/twist-cli/SKILL.md.cursor/skills/twist-cli/SKILL.md.gemini/skills/twist-cli/SKILL.md.pi/skills/twist-cli/SKILL.md.agents/skills/twist-cli/SKILL.md

Decision evidence

public snapshot

AI identified this as a dangerous AI-agent capability at 96.0% confidence with low false-positive risk.

Evidence for policy risk

The package registers a postinstall lifecycle hook.
Postinstall imports code that updates all globally installed skills.
The update loop overwrites every detected installed skill without a user command.
Targeted skill locations include Claude Code, Codex, Cursor, Gemini, Pi, and the broad .agents directory.
Skill updates write a package-controlled SKILL.md into those agent control surfaces.
The updater overwrites the installed SKILL.md with package-controlled content.

Evidence against

The updater skips agent skill paths that are not already present.
No relevant network endpoint or remote payload loader was found in the install-time skill-update path.
The interactive skill commands expose explicit install, update, and uninstall actions.

Behavioral surface

Source

ChildProcessDynamicRequireEnvironmentVarsFilesystemNetworkShell

Supply chain

HighEntropyStrings

Source & flagged code

7 flagged

package.jsonView file

Lines 1-37json

2 "name": "@doist/twist-cli",

3 "version": "3.0.8",

4 "description": "TypeScript CLI for Twist",

5 "type": "module",

6 "main": "dist/index.js",

7 "bin": {

8 "tw": "dist/index.js"

9 },

10 "scripts": {

11 "build": "tsc -p tsconfig.build.json && chmod +x dist/index.js",

12 "dev": "tsc -p tsconfig.build.json --watch",

13 "start": "node dist/index.js",

14 "type-check": "tsc --noEmit",

15 "lint": "oxlint . --fix && oxfmt",

16 "lint:check": "oxlint . && oxfmt --check",

17 "postinstall": "node scripts/postinstall.js",

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg · L17

Medium

Ambiguous Install Lifecycle Script

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

package.jsonView on unpkg · L17

18 "check:skill-sync": "node scripts/check-skill-sync.js",

19 "sync:skill": "npm run build && node scripts/sync-skill.js",

20 "test": "vitest run",

21 "test:watch": "vitest",

22 "prepublishOnly": "npm run build && npm test"

23 },

24 "keywords": [

25 "twist",

26 "cli",

27 "messaging",

28 "doist"

29 ],

30 "author": "Ernesto García <ernesto@ernesto.dev>",

31 "license": "MIT",

32 "engines": {

33 "node": ">=24",

34 "npm": ">=11"

35 },

36 "repository": {

37 "type": "git",

•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 1

High

Semantic Analysis Limited

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

package.jsonView on unpkg

scripts/check-skill-sync.jsView file

Lines 1-37javascript

1#!/usr/bin/env node

3/**

4 * Verifies that skills/twist-cli/SKILL.md is in sync with the generated

5 * skill content from src/lib/skills/content.ts.

6 *

7 * Requires `npm run build` to have been run first.

8 */

10import { readFile } from 'node:fs/promises'

dist/lib/skills/create-installer.jsView file

Lines 4-44javascript

4import { CliError } from '../errors.js';

5import { SKILL_FILE_CONTENT, SKILL_NAME } from './content.js';

6export function createInstaller(config) {

7 function getInstallPath

dist/lib/input.jsView file

•matchType = normalized_sha256 matchedPackage = @doist/twist-cli@3.0.7 matchedPath = dist/lib/input.js matchedIdentity = npm:QGRvaXN0L3R3aXN0LWNsaQ:3.0.7 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.

dist/lib/input.jsView on unpkg

•matchType = malicious_source_fingerprint_signature signature = acd00388ab62c770 signatureType = suspicious_hashes sourceLabel = final_verdict:malicious matchedPackage = @doist/twist-cli@3.0.7 matchedPath = dist/lib/input.js matchedIdentity = npm:QGRvaXN0L3R3aXN0LWNsaQ:3.0.7 similarity = 1.000 shingleOverlap = 22 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.

dist/lib/input.jsView on unpkg

Source & flagged code

7 flagged · loading source

package.jsonView file

•scripts.postinstall = node scripts/postinstall.js

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg

•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 1

High

Semantic Analysis Limited

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

package.jsonView on unpkg

•scripts.postinstall = node scripts/postinstall.js

Medium

Ambiguous Install Lifecycle Script

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

package.jsonView on unpkg

scripts/check-skill-sync.jsView file

Findings

1 Critical4 High5 Medium5 Low

CriticalAi Agent Control Hijackdist/lib/skills/create-installer.js

HighInstall Time Lifecycle Scriptspackage.json

HighSemantic Analysis Limitedpackage.json

HighKnown Malware Source Similaritydist/lib/input.js

HighKnown Malware Source Fingerprint Signaturedist/lib/input.js

MediumAmbiguous Install Lifecycle Scriptpackage.json

MediumDynamic Requirescripts/check-skill-sync.js

MediumNetwork

MediumEnvironment Vars

MediumStructural Risk Force Deep Review

LowNon Install Lifecycle Scripts

LowScripts Present

LowFilesystem

Low

Affected versions and remediation

This report applies to @doist/twist-cli@3.0.8.

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

Avoid installing @doist/twist-cli@3.0.8. 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: 2026-09-11 03:40:39.614Z (UTC).

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

Versions

14 known versions

3.0.15latestSuspiciousAgent extension lifecycle riskWarnPublishedOct 19, 2026, 12:27 PM 3.0.14SuspiciousAgent extension lifecycle riskWarnPublishedOct 12, 2026, 4:14 PM 3.0.13SuspiciousAgent extension lifecycle riskWarnPublishedSep 28, 2026, 4:31 AM 3.0.12SuspiciousAgent extension lifecycle riskWarnPublishedSep 21, 2026, 11:02 PM

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Affected versions and remediation

Versions

Related packages

More from @doist

Other AI-agent control-surface risk packages

Behavioral surface

Related packages

More from @doist

Other AI-agent control-surface risk packages

Source & flagged code