registry / AI-agent capability abuse / alvin-bot / 5.103.0

alvin-bot@5.103.0

Self-hosted autonomous AI agent on Telegram, Slack, Discord, WhatsApp, Signal, terminal & web. Claude Agent SDK + multi-provider engine with auto-failover, detached sub-agents, zero-config memory. Local-first, telemetry-free.

AI Security Review

scanned 1mo ago · by lpm-firewall-ai

The explicit permission bypass and prompt scrubbing warrant a warning for dangerous agent capability. Static inspection did not establish a concrete malicious payload or covert data export.

Static reason

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

Trigger

A user starts the bot and a request is handled through the Claude bridge.

Impact

A running agent can operate without the normal Claude permission confirmation control, increasing the risk of unintended tool actions.

Mechanism

The bridge scrubs prompt identifiers, then spawns the Claude command with the dangerously-skip-permissions option.

AI rationale

The explicit permission bypass and prompt scrubbing warrant a warning for dangerous agent capability. Static inspection did not establish a concrete malicious payload or covert data export.

Evidence

claude-bridge/src/claude.js

Decision evidence

public snapshot

AI called this Suspicious at 90.0% confidence as Dangerous Capability with low false-positive risk.

Evidence for warning

The package exposes a user-run command-line entry point.
The Claude bridge rewrites OpenClaw identifiers before forwarding its system prompt.
The bridge starts Claude with its permission checks explicitly disabled.

Evidence against

package.json contains no preinstall, install, or postinstall lifecycle hook.
No inspected code showed secret collection or transmission to a package-controlled endpoint.
The only identified credential check sends a supplied Telegram token to Telegram's own API.

Behavioral surface

Source

ChildProcessCryptoDynamicRequireEnvironmentVarsEvalFilesystemNetworkShellWebSocket

Supply chain

HighEntropyStringsMinifiedObfuscatedUrlStrings

Manifest

NoLicense

scanned 278 file(s), 3.89 MB of source, external domains: 127.0.0.1, ai.google.dev, aistudio.google.com, api.anthropic.com, api.coingecko.com, api.frankfurter.app, api.github.com, api.groq.com, api.openai.com, api.telegram.org, brew.sh, build.nvidia.com, cdn.jsdelivr.net, chat.openai.com, claude.ai, console.anthropic.com, console.groq.com, discord.com, generativelanguage.googleapis.com, integrate.api.nvidia.com, ollama.com, openrouter.ai, platform.openai.com, query2.finance.yahoo.com, t.me, wacli.sh, wttr.in, www.apple.com

Source & flagged code

20 flagged

claude-bridge/src/claude.jsView file

Lines 1-22javascript

1'use strict';

3const { spawn } = require('child_process');

4const crypto = require('crypto');

High

Child Process

Package source references child process execution.

claude-bridge/src/claude.jsView on unpkg · L2

Medium

Dynamic Require

Package source references dynamic require/import behavior.

claude-bridge/src/claude.jsView on unpkg · L2

6const CLAUDE_BIN = process.env.CLAUDE_BIN || 'claude';

Low

Weak Crypto

Package source references weak cryptographic algorithms.

claude-bridge/src/claude.jsView on unpkg · L2

8// --- Stable alias per CLI session ---

9// Each session keeps one alias so resumed requests don't accumulate different

10// names in the Claude context (which would leak un-replaced aliases in output).

11const PREFIXES = ['Chat', 'Dev', 'Run', 'Ask', 'Net', 'App', 'Zen', 'Arc', 'Dot', 'Amp', 'Hex', 'Orb', 'Elm', 'Oak', 'Sky'];

12const SUFFIXES = ['Kit', 'Box', 'Pod', 'Hub', 'Lab', 'Ops', 'Bay', 'Tap', 'Rim', 'Fog', 'Dew', 'Fin', 'Gem', 'Jet', 'Cog'];

13const pick = (arr) => arr[Math.floor(Math.random() * arr.length)];

14const sessionAliasMap = new Map(); // sessionId → { alias, aliasLower, lastUsed }

15const sessionTokenMaps = new Map(); // sessionId → Map<original, replacement>

17function getSessionAlias(sessionId) {

18 if (!sessionId) {

19 const alias = pick(PREFIXES) + pick(SUFFIXES);

20 return { alias, aliasLower: alias.toLowerCase() };

21 }

22 let entry = sessionAliasMap.get(sessionId);

•matchType = normalized_sha256 matchedPackage = alvin-bot@5.149.0 matchedPath = claude-bridge/src/claude.js matchedIdentity = npm:YWx2aW4tYm90:5.149.0 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.

claude-bridge/src/claude.jsView on unpkg

bin/cli.jsView file

22Manifest entrypoint (manifest.bin) carries capability families absent from dist/build output: environment+network, sensitive-file+network, execution+network L22: import { homedir } from "os"; L23: import { execSync } from "child_process"; L24: import { initI18n, t, getLocale } from "../dist/i18n.js"; ... L28: // Data directory — same logic as src/paths.ts L29: const DATA_DIR = process.env.ALVIN_DATA_DIR || join(homedir(), ".alvin-bot"); L30: ... L42: const ensureRL = () => { L43: if (!rl) rl = createInterface({ input: process.stdin, output: process.stdout }); L44: return rl; ... L83: try { L84: return JSON.parse(readFileSync(resolve(import.meta.dirname || ".", "../package.json"), "utf-8")).version; L85: } catch { return ""; }

High

Entrypoint Build Divergence

Manifest entrypoint contains risky behavior absent from dist/build output.

bin/cli.jsView on unpkg · L22

1312console.log(`\n The Claude Agent SDK needs the native Claude Code binary.`); L1313: console.log(` Install with:\n curl -fsSL https://claude.ai/install.sh | sh`); L1314: console.log(` (npm install @anthropic-ai/claude-code is the alternative path.)\n`); ... L1320: try { L1321: execSync("curl -fsSL https://claude.ai/install.sh | sh", { stdio: "inherit", timeout: 120_000 }); L1322: const localBin = join(homedir(), ".local", "bin"); L1323: if (!process.env.PATH.includes(localBin)) { L1324: process.env.PATH = `${localBin}:${process.env.PATH || ""}`;

High

Same File Env Network Execution

A single source file combines environment access, network access, and code or shell execution; review context before blocking.

bin/cli.jsView on unpkg · L1312

22import { homedir } from "os"; L23: import { execSync } from "child_process"; L24: import { initI18n, t, getLocale } from "../dist/i18n.js"; ... L28: // Data directory — same logic as src/paths.ts L29: const DATA_DIR = process.env.ALVIN_DATA_DIR || join(homedir(), ".alvin-bot"); L30: ... L42: const ensureRL = () => { L43: if (!rl) rl = createInterface({ input: process.stdin, output: process.stdout }); L44: return rl; ... L83: try { L84: return JSON.parse(readFileSync(resolve(import.meta.dirname || ".", "../package.json"), "utf-8")).version; L85: } catch { return ""; }

dist/web/server.jsView file

Lines 1-1javascript

1const _0xda9014=_0x5717,_0x109502=_0x5717;(function(_0x2d7adc,_0x4f52a6){const _0x58958d=_0x5717,_0x566d62=_0x5717,_0x3a2e97=_0x2d7adc();while(!!

dist/middleware/auth.jsView file

1function _0x5df0(_0x22646e,_0xc771dd){_0x22646e=_0x22646e-(0x1*-0x1f5b+-0x1*-0x94e+0x6*0x3ef);const _0x573403=_0x2cd6();let _0x413add=_0x573403[_0x22646e];if(_0x5df0['KRgino']===un...

High

Obfuscated Payload Loader

Source contains an obfuscated payload loader that reconstructs and executes hidden code.

dist/middleware/auth.jsView on unpkg · L1

install.shView file

•path = install.sh kind = build_helper sizeBytes = 10656 magicHex = [redacted]

Medium

Ships Build Helper

Package ships non-JavaScript build or shell helper files.

install.shView on unpkg

claude-bridge/src/claude.js#virtual:normalized:round1View file

•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 17

High

Semantic Analysis Limited

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

claude-bridge/src/claude.js#virtual:normalized:round1View on unpkg

claude-bridge/src/server.jsView file

•matchType = normalized_sha256 matchedPackage = alvin-bot@5.149.0 matchedPath = claude-bridge/src/server.js matchedIdentity = npm:YWx2aW4tYm90:5.149.0 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.

claude-bridge/src/server.jsView on unpkg

plugins/email/index.jsView file

•matchType = normalized_sha256 matchedPackage = alvin-bot@5.149.0 matchedPath = plugins/email/index.js matchedIdentity = npm:YWx2aW4tYm90:5.149.0 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.

plugins/email/index.jsView on unpkg

Source & flagged code

20 flagged · loading source

claude-bridge/src/claude.jsView file

2L3: const { spawn } = require('child_process'); L4: const crypto = require('crypto');

High

Child Process

Package source references child process execution.

claude-bridge/src/claude.jsView on unpkg · L2

High

Known Malware Source Similarity

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

claude-bridge/src/claude.jsView on unpkg

2L3: const { spawn } = require('child_process'); L4: const crypto = require('crypto');

Medium

Dynamic Require

Package source references dynamic require/import behavior.

claude-bridge/src/claude.jsView on unpkg · L2

Findings

17 High6 Medium8 Low

HighChild Processclaude-bridge/src/claude.js

HighShell

HighEntrypoint Build Divergencebin/cli.js

HighSame File Env Network Executionbin/cli.js

HighCommand Output Exfiltrationdist/web/server.js

HighSandbox Evasion Gated Capabilitybin/cli.js

HighObfuscated Payload Loaderdist/middleware/auth.js

HighCross File Remote Execution Contextbin/cli.js

HighTrigger Reachable Command Output Exfiltrationdist/web/server.js

HighTrigger Reachable Persistencebin/cli.js

HighRuntime Package Installbin/cli.js

HighObfuscated

HighSemantic Analysis Limited

Affected versions and remediation

This report applies to alvin-bot@5.103.0.

See all versions of alvin-bot for other recorded verdicts.

Review the evidence and your use of alvin-bot@5.103.0 before allowing it. Restrict the permissions described in this report.
Choose an independently verified alternative or release. This report does not establish that other versions are safe.

Evidence last updated: 2026-09-22 05:00:39.609Z (UTC).

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

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Affected versions and remediation

Runtime surface

Behavioral surface

Source & flagged code