registry / @cakagus/codeclaw / 1.1.1
@cakagus/codeclaw@1.1.1
AI agent terminal + Telegram bot untuk coding, trading, dan research — dengan skill system auto-upgrade.
AI Security Review
scanned 5d ago · by lpm-firewall-aiThe critical routing hint is explained by requests that use configured keys with their corresponding services and an explicitly configured Telegram destination. The package has substantial network and execution capabilities, but the inspected behavior matches its documented agent, research, trading, and bot functions and did not reveal an unrelated or automatic attack.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
User starts the CLI, runs the market-analysis script, or submits a voice recording to the configured Telegram bot.
Impact
Expected service requests and report delivery; no unrelated credential collection or unauthorized account action was identified.
Mechanism
The package sends requests to configured model, market-data, search, Telegram, and transcription services; the flagged market-analysis path sends its report to the configured Telegram chat.
AI rationale
The critical routing hint is explained by requests that use configured keys with their corresponding services and an explicitly configured Telegram destination. The package has substantial network and execution capabilities, but the inspected behavior matches its documented agent, research, trading, and bot functions and did not reveal an unrelated or automatic attack.
Evidence
config.jsongold.logtelegram.pid.healthdownloads/
Network endpoints6
api.exa.ai/contentsapi.exa.ai/searchapi.exa.ai/embeddingsapi.twelvedata.com/time_seriesapi.telegram.orgapi.groq.com/openai/v1/audio/transcriptions
Decision evidence
public snapshotAI called this Clean at 91.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The manifest defines ordinary start and Telegram scripts, with no install lifecycle hook.
- The Exa API key is sent to api.exa.ai for Exa content requests.
- The gold-analysis script reads service keys and Telegram destination from config.json.
- Telegram reports are sent only when the configured bot token and chat ID are present; the script runs as a command and supports --no-telegram.
- Audio transcription sends a configured Groq key and a user-provided voice recording to Groq’s transcription endpoint.
Evidence against
- The package provides network, filesystem, shell, and deployment capabilities, but the inspected examples are tied to its documented agent, Telegram, research, and trading features.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsUrlStrings
Source & flagged code
4 flaggedbin/codeclaw.jsView file
31case "tg":
L32: await import(path.join(PKG_ROOT, "telegram.js"));
L33: break;
Medium
Dynamic Require
Package source references dynamic require/import behavior.
bin/codeclaw.jsView on unpkg · L31scripts/gold-analysis.mjs#virtual:normalized:round1View file
55try {
L56: const res = await fetch("https://nfs.faireconomy.media/ff_calendar_thisweek.json", {
L57: signal: AbortSignal.timeout(10000),
...
L59: if (!res.ok) return null;
L60: const data = await res.json();
L61: const today = new Date().toISOString().slice(0, 10);
...
L230: if (data?.status === "error" || data?.code) {
L231: throw new Error(`Twelve Data: ${data.message || `code ${data.code}`}`);
L232: }
Critical
Hardcoded Runtime Data Exfiltration
Source sends credentials or rich application records to a package-controlled external receiver enabled by default.
scripts/gold-analysis.mjs#virtual:normalized:round1View on unpkg · L55Findings
1 Critical1 High5 Medium5 Low
CriticalHardcoded Runtime Data Exfiltrationscripts/gold-analysis.mjs#virtual:normalized:round1
HighSemantic Analysis Limitedlib/research.js#virtual:normalized:round1
MediumDynamic Requirebin/codeclaw.js
MediumNetwork
MediumEnvironment Vars
MediumShips Build Helperbin/tokenobral-deploy.sh
MediumStructural Risk Force Deep Review
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
LowNo License
Affected versions and remediation
This report applies to @cakagus/codeclaw@1.1.1.
See all versions of @cakagus/codeclaw for other recorded verdicts.
- Review the evidence and your use of @cakagus/codeclaw@1.1.1 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.