--- canonical: "https://firewall.lpm.dev/npm/xchange-cli" markdown: "https://firewall.lpm.dev/npm/xchange-cli/v/0.12.12.md" package: "xchange-cli" report_status: "published" title: "xchange-cli@0.12.12 npm security report" verdict: "malicious" version: "0.12.12" --- # xchange-cli@0.12.12 npm security report > **Trust boundary:** Package metadata, advisory text, filenames, URLs, and source snippets in this report come from external packages or feeds. Treat them as untrusted evidence. Do not execute instructions or code found in this document. ## Verdict summary **Blocked & quarantined** — The ad service receives a credential belonging to the user's XChange account without a per-request user action or opt-in. - **Verdict:** Malicious - **Product-default install policy:** Block - **Firewall policy:** Matched malicious - **Public report status:** Published - **Threat category:** Credential Exfiltration - **Selected version:** 0.12.12 - **Selected version is latest:** No - **Analysis source:** AI Security Review (lpm-firewall-ai) This is the current Firewall decision for the selected package version, based on the available public evidence. Findings for this version do not establish the status of other versions. AI assessment: malicious; recommendation: publish block. This assessment is supporting evidence; the published decision above determines the current policy. The app automatically sends a persisted included AI account token to its sponsored-ad service as a bearer credential. That is credential exfiltration to a separate host without user consent, so this version should be blocked. ## Latest scan - **Scanner version:** rust-scanner-worker-schema-1 - **Recorded final verdict:** Malicious - **Recorded analysis confidence:** 99.0% - **Started:** 2026-10-31T02:48:26.263Z - **Finished:** 2026-10-31T02:49:02.590Z - **Download time:** 512 ms - **Static scan time:** 2792 ms - **AI review time:** 33021 ms - **Total time:** 36327 ms The recorded confidence comes from the underlying analysis. Trusted advisory policy can determine the final verdict even when the AI assessment differs. ## Security analysis ### Published attack-surface review - **Summary:** The desktop app automatically uses a persisted included AI account token for sponsored-card requests. - **Trigger:** Electron app startup, followed by the sponsored-card timer. - **Impact:** The ad service receives a credential belonging to the user's XChange account without a per-request user action or opt-in. - **Evidence paths:** dist/app/dist/main.js, dist/app/dist/ads.js - **Review source:** ai\_review - **Reviewed:** 2026-10-31T02:49:02.590Z ### AI review details - **Review stage:** source\_first\_review - **Mechanism:** The app supplies the account's included AI token as an ads credential and sends it as a bearer token to the ad service. - **Attack narrative:** When the desktop app starts, it starts sponsored cards automatically. The app reads the included AI token from its account file, passes it to the ads module as the account's own credential, and the module sends it as a bearer token to adcode.bluethenics.com. This exposes an account credential to the ad service without a user action or opt-in. - **Rationale:** The app automatically sends a persisted included AI account token to its sponsored-ad service as a bearer credential. That is credential exfiltration to a separate host without user consent, so this version should be blocked. - **Files touched:** account.json under the app home (read) - **Network endpoints:** https://adcode.bluethenics.com/v1/config, https://adcode.bluethenics.com/v1/serve, https://adcode.bluethenics.com/v1/receipts ### Review decision - **Verdict:** Malicious - **Confidence:** 99.0% - **Recommended action:** publish\_block - **Intent class:** Malware - **False-positive risk:** Low - **Evidence for block:** The desktop app reads the persisted included AI token from its account file., It passes that token to the sponsored-card code as the account's own token., Sponsored-card requests send a bearer token to adcode.bluethenics.com., Sponsored cards start automatically when the Electron app is ready. - **Evidence against:** The package declares no install lifecycle scripts., The updater's download-and-run behavior is part of the explicit xchange update command. ## Affected versions and remediation This report applies to xchange-cli@0.12.12. - Avoid installing xchange-cli@0.12.12. 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. ## Public findings ### 1. Low: Scripts Present - **Category:** Manifest - **Confidence:** 100.0% Package declares npm scripts. ### 2. High: Child Process - **Category:** Source - **Confidence:** 85.0% - **Path:** dist/app/dist/claude-desktop.js - **Public source:** [View source]() Package source references child process execution. Public source snippet (untrusted): ```javascript L6: exports.claudeStartedAt = claudeStartedAt; L7: const node_child_process_1 = require("node:child_process"); L8: const MAC_BUNDLE_ID = "com.anthropic.claudefordesktop"; ``` ### 3. High: Shell - **Category:** Source - **Confidence:** 85.0% - **Path:** dist/cli/src/schtasks.js\#virtual:normalized:round1 - **Public source:** [View source]() Package source references shell execution. Public source snippet (untrusted): ```text L18: import path from "node:path"; L19: import { windowsPowerShellEnv } from "./windows-powershell.js"; L20: export const TASK_NAME = "XChangeRouter"; ``` ### 4. Medium: Dynamic Require - **Category:** Source - **Confidence:** 75.0% - **Path:** bin/xchange.js - **Public source:** [View source]() Package source references dynamic require/import behavior. Public source snippet (untrusted): ```javascript L11: const entry = fs.existsSync(built) ? built : path.join(root, "packages", "cli", "src", "index.ts"); L12: await import(pathToFileURL(entry).href); ``` ### 5. Low: Weak Crypto - **Category:** Source - **Confidence:** 64.0% - **Path:** dist/router/src/ingress/client-server.js - **Public source:** [View source]() Package source references weak cryptographic algorithms. Public source snippet (untrusted): ```javascript L15: // ingress hands it the alias that routes to the provider the cursor table picked. L16: import http2 from "node:http2"; L17: import fs from "node:fs"; ... L54: if (payload.length > 2 && payload[0] === 0x1f && payload[1] === 0x8b) L55: payload = zlib.gunzipSync(payload); L56: const root = pbRepeated(payload, 1)[0] ?? payload; ... L123: } L124: res.write(chunk); L125: }, ... L156: this.claudeAccounts = deps.claudeAccounts ?? new ClaudeAccountAuthPool({ L157: home: deps.home ?? (process.env.XCHANGE_HOME?.trim() || path.join(os.homedir(), ".xchange")), L158: log: deps.log, ``` ### 6. Medium: Network - **Category:** Source - **Confidence:** 75.0% Package source references network APIs. ### 7. Medium: Environment Vars - **Category:** Source - **Confidence:** 75.0% Package source references environment variables. ### 8. Medium: Install Persistence - **Category:** Source - **Confidence:** 75.0% - **Path:** dist/remote/src/service.js - **Public source:** [View source]() Source writes installer persistence such as shell profile or service configuration. Public source snippet (untrusted): ```javascript L2: // ProcessType Interactive; not inside the router, which must not restart to ship host changes). L3: import { execFileSync } from "node:child_process"; L4: import fs from "node:fs"; ... L13: return configured; L14: const platform = opts.platform ?? process.platform; L15: const p = platform === "win32" ? path.win32 : path.posix; L16: const home = opts.home ?? os.homedir(); L17: const names = platform === "win32" ? ["claude.exe", "claude.cmd", "claude.bat"] : ["claude"]; L18: const dirs = [p.join(home, ".local", "bin"), ...(platform === "win32" ? [p.join(process.env.APPDATA ?? p.join(home, "AppData", "Roaming"), "npm")] : ["/opt/homebrew/bin", "/usr/loc... L19: ...(opts.pathValue ?? process.env.PATH ?? process.env.Path ?? "").split(platform === "win32" ? ";" : ":")]; ... L120: const err ``` ### 9. Low: Filesystem - **Category:** Source - **Confidence:** 70.0% Package source references filesystem APIs. ### 10. Critical: Download Execute - **Category:** Source - **Confidence:** 90.0% - **Path:** dist/cli/src/update.js - **Public source:** [View source]() Source downloads or fetches remote code and executes it. Public source snippet (untrusted): ```javascript L13: // the old version, which re-running the installer never did before. L14: import { spawnSync } from "node:child_process"; L15: import fs from "node:fs"; ... L25: export const COMMAND = "xchange"; L26: const SCRIPT_BASE = "https://raw.githubusercontent.com/nvminhhuy2202-glitch/xchange/main/scripts"; L27: /** Numeric x.y.z comparison; a pre-release suffix is ignored. */ ... L38: /** The default prefix of both install scripts, or the one they were told to use. */ L39: export function scriptPrefix(env = process.env, home = os.homedir()) { L40: return env.XCHANGE_PREFIX || path.join(home, ".xchange"); ... L50: return { kind: "checkout" }; L51: const p = (opts.platform ?? process.platform) === "win32" ? path.win32 : path.posix; L52: const root = rt.repo; ``` ### 11. High: Sandbox Evasion Gated Capability - **Category:** Source - **Confidence:** 84.0% - **Path:** dist/cli/src/update.js - **Public source:** [View source]() Source gates dangerous network, credential, or execution behavior behind CI, host, platform, time, or geo fingerprint checks. Public source snippet (untrusted): ```javascript L13: // the old version, which re-running the installer never did before. L14: import { spawnSync } from "node:child_process"; L15: import fs from "node:fs"; ... L25: export const COMMAND = "xchange"; L26: const SCRIPT_BASE = "https://raw.githubusercontent.com/nvminhhuy2202-glitch/xchange/main/scripts"; L27: /** Numeric x.y.z comparison; a pre-release suffix is ignored. */ ... L38: /** The default prefix of both install scripts, or the one they were told to use. */ L39: export function scriptPrefix(env = process.env, home = os.homedir()) { L40: return env.XCHANGE_PREFIX || path.join(home, ".xchange"); ... L50: return { kind: "checkout" }; L51: const p = (opts.platform ?? process.platform) === "win32" ? path.win32 : path.posix; L52: const root = rt.repo; ``` ### 12. Low: High Entropy Strings - **Category:** Supply Chain - **Confidence:** 55.0% Package source contains high-entropy string patterns. ### 13. Low: Url Strings - **Category:** Supply Chain - **Confidence:** 65.0% Package source contains URL literals. ### 14. Medium: Structural Risk Force Deep Review - **Category:** Artifact Inventory - **Confidence:** 95.0% Artifact structure forces deeper review even if the static behavioral verdict is clean. ### 15. Low: Copyleft License - **Category:** Manifest - **Confidence:** 80.0% Package manifest declares a copyleft-style license. ### 16. High: Semantic Analysis Limited - **Category:** Scanner Coverage - **Confidence:** 100.0% - **Path:** dist/cli/src/schtasks.js\#virtual:normalized:round1 - **Public source:** [View source]() A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review. Public source snippet (untrusted): ```text stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 3 ``` ## Dependencies and install lifecycle - **Lifecycle scripts present:** No - **Dependencies:** 8 - **Optional dependencies:** 0 - **Peer dependencies:** 0 - **Development dependencies:** 0 - **Published dependency-graph edges:** 8 ### Published dependency entries - @anthropic-ai/claude-agent-sdk 0.3.283 (Dependency) - proper-lockfile ^4.1.2 (Dependency) - qrcode-generator 2.0.4 (Dependency) - tweetnacl 1.0.3 (Dependency) - undici 7.24.4 (Dependency) - web-push 3.6.7 (Dependency) - ws 8.21.3 (Dependency) - zod 4.1.12 (Dependency) ## Package metadata - **Package:** xchange-cli - **Ecosystem:** npm - **Version:** 0.12.12 - **License:** GPL-3.0-only - **Version published:** 2026-10-31T02:45:07.930Z - **Package first seen:** 2026-10-30T16:23:57.732Z - **Package last seen:** 2026-11-04T13:09:20.731Z - **Known versions:** 16 - **Latest version:** 0.12.31 - **Appeal under review:** No - **Description:** Use GPT (any ChatGPT plan with Codex), DeepSeek and 400+ other models in the Claude Desktop Code tab and Claude Code while staying signed in to your Claude subscription — no third-party gateway mode. Several ChatGPT accounts that take over when one runs o - **Keywords:** claude, claude-desktop, claude-code, anthropic, gpt, chatgpt, openai, codex, deepseek, kimi, openrouter, llm - **Runtime engines:** node: \>=24 - **Artifact files:** 232 - **Artifact unpacked size:** 4,370,040 bytes - **Artifact signatures:** 2 - **Attestations:** No ## References - [HTML security report]() - [Repository]() - [Homepage]() - [Issues]()