codebolt@1.12.88
CodeBolt headless server CLI
AI Security Review
scanned 4d ago · by lpm-firewall-aiThe server configures outbound logging to a vendor endpoint by default, while the visible opt-out applies only to PostHog. Because the logged payload is not established here, this is suspicious runtime telemetry rather than confirmed exfiltration.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
A user starts the CLI or server.
Impact
Runtime logging may disclose identifying or sensitive details without a clear consent control, so this is a telemetry concern rather than a confirmed data-exfiltration attack.
Mechanism
The server initializes a categorized logger configured for Cloudflare delivery, with a fixed endpoint and token; only PostHog is shown with an opt-out.
AI rationale
The server configures outbound logging to a vendor endpoint by default, while the visible opt-out applies only to PostHog. Because the logged payload is not established here, this is suspicious runtime telemetry rather than confirmed exfiltration.
Evidence
dist/index.js
Network endpoints1
codebolt-log-ingest.arrowai.workers.dev
Decision evidence
public snapshotAI called this Suspicious at 82.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- Server startup configures structured logging to send to https://codebolt-log-ingest.arrowai.workers.dev with Cloudflare enabled by default; the shown opt-out controls PostHog only.
- The CLI launcher requires dist/index.js when a user runs the command.
- The manifest lists a prepublishOnly build script, not an install, preinstall, or postinstall hook.
Evidence against
- The outbound logging is part of the package’s server runtime, and the inspected source does not establish that it transmits credentials, prompts, code, or tool output.
- The scanner’s download-and-run lead resolves to a bundled Bash syntax definition rather than an executed download command.
Behavioral surface
CryptoEnvironmentVarsEvalFilesystemNetworkWebSocket
HighEntropyStringsMinifiedPossibleObfuscationTelemetryUrlStrings
Source & flagged code
7 flaggeddist/web/assets/serverWorkerMain-CPdrTHxh.jsView file
2patternName = generic_password
severity = medium
line = 2
Medium
Secret Pattern
Package contains a possible secret pattern.
dist/web/assets/serverWorkerMain-CPdrTHxh.jsView on unpkg · L21/*! For license information please see serverWorkerMain.js.LICENSE.txt */
L2: var serverExportVar;(()=>{var e={3608:(e,t,r)=>{"use strict";var n;function i(e){if("contents"in e)n="string"==typeof e.contents?JSON.parse(e.contents):e.contents;else if("fsPath"i...
Low
Eval
Package source references a known benign dynamic code generation pattern.
dist/web/assets/serverWorkerMain-CPdrTHxh.jsView on unpkg · L1dist/web/assets/iconv-lite-umd-DpoLQuPU.jsView file
7contains invisible/control Unicode U+200B (zero width space)
\v\f\r\x1B !"#$٪&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_\`abcdefghijklmnopqrstuvwxyz{|}~°·∙√▒─│┼┤┬├┴┐┌└┘β∞φ±½¼≈«»ﻷﻸ��ﻻﻼ� ﺂ£¤ﺄ��ﺎﺏﺕﺙ،ﺝﺡﺥ٠١٢٣٤٥٦٧٨٩ﻑ؛ﺱﺵﺹ؟¢ﺀﺁﺃﺅﻊﺋﺍﺑﺓﺗﺛﺟﺣﺧﺩﺫﺭﺯﺳﺷﺻﺿﻁﻅﻋﻏ¦¬÷×ﻉـﻓﻗﻛﻟﻣﻧﻫﻭﻯﻳﺽﻌﻎﻍﻡﹽّﻥ
Findings
1 Critical2 High7 Medium8 Low
CriticalTrojan Source Unicodedist/web/assets/iconv-lite-umd-DpoLQuPU.js
HighCross File Remote Execution Contextdist/web/assets/htmlServerMain-BEIBCfWH.js
HighSemantic Analysis Limiteddist/web/assets/elk.bundled-DED6nYWl.js
MediumSecret Patterndist/web/assets/serverWorkerMain-CPdrTHxh.js
MediumNetwork
MediumEnvironment Vars
MediumShips Wasm Moduledist/web/assets/onig-Du5pRr7Y.wasm
MediumOversized Source Filedist/web/assets/vendor-monaco-7uIto7rp.js
MediumStructural Risk Force Deep Review
MediumWildcard Dependency
LowNon Install Lifecycle Scripts
LowScripts Present
LowEval
Affected versions and remediation
This report applies to codebolt@1.12.88.
See all versions of codebolt for other recorded verdicts.
- Review the evidence and your use of codebolt@1.12.88 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: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.