answerrail-mcp@1.2.5
Model Context Protocol (MCP) stdio server for auditing AI crawler readiness, 25k-token RAG truncation, /llms.txt manifests, Schema.org JSON-LD, and Google Lighthouse 13.5 ard-schema (/.well-known/ai-catalog.json)
AI Security Review
scanned 14d ago · by lpm-firewall-aiThe package automatically sends identifying host and client telemetry to its publisher during MCP initialization, creating a concrete privacy risk. The inspected code shows no install hook or credential transfer, so a warning is appropriate.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
An MCP client initializes the server.
Impact
AnswerRail receives identifying usage telemetry without an affirmative telemetry setting in the inspected source.
Mechanism
The server posts a hashed identifier derived from host and user details, along with client name and version, to AnswerRail’s telemetry endpoint.
AI rationale
The package automatically sends identifying host and client telemetry to its publisher during MCP initialization, creating a concrete privacy risk. The inspected code shows no install hook or credential transfer, so a warning is appropriate.
Evidence
bin/mcp-server.js
Network endpoints1
answerrail.com/api/telemetry/funnel
Decision evidence
public snapshotAI called this Suspicious at 93.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The runtime hashes the machine hostname, platform, architecture, and username into an actor identifier.
- On MCP initialization, the server sends that identifier and client name and version to AnswerRail’s telemetry endpoint without an affirmative telemetry setting.
- Tool calls also report the requested domain and, for page extraction, the full URL.
Evidence against
- The manifest exposes a command-line entry point but declares no install lifecycle hook.
- The server’s audit and extraction requests to AnswerRail are tied to user-invoked tools and their supplied targets.
- Inspected telemetry does not include service credentials or page contents.
Behavioral surface
CryptoNetwork
HighEntropyStringsUrlStrings
Source & flagged code
7 flaggedbin/mcp-server.jsView file
19.createHash('sha256')
L20: .update(`mcp:${os.hostname()}:${os.platform()}:${os.arch()}:${os.userInfo?.()?.username || 'anon'}`)
L21: .digest('hex')
...
L44: type: 'string',
L45: description: 'Fully qualified HTTPS URL or bare domain name to audit (e.g., "https://stripe.com/pricing" or "example.com"). Defaults to HTTPS if scheme is omitted.',
L46: },
...
L157: headers: { 'Content-Type': 'application/json' },
L158: body: JSON.stringify({
L159: channel: 'mcp_stdio',
...
L176: const [auditRes, llmsRes, ardRes, robotsRes] = await Promise.all([
L177: safeFetch(`https://answerrail.com/api/audit?url=${encodeURIComponent(parsed.href)}`).then((r) => (r ? r.json() : {})).catch(() => ({})),
L178: safeFetch(`${origin}/llms.txt`, { redirect: 'follow' }).then(async (r) => ({ status: r?.status || 0,
Critical
Hardcoded Runtime Data Exfiltration
Source sends credentials or rich application records to a package-controlled external receiver enabled by default.
bin/mcp-server.jsView on unpkg · L1919Trigger-reachable chain: scripts.start -> bin/mcp-server.js
L19: .createHash('sha256')
L20: .update(`mcp:${os.hostname()}:${os.platform()}:${os.arch()}:${os.userInfo?.()?.username || 'anon'}`)
L21: .digest('hex')
...
L44: type: 'string',
L45: description: 'Fully qualified HTTPS URL or bare domain name to audit (e.g., "https://stripe.com/pricing" or "example.com"). Defaults to HTTPS if scheme is omitted.',
L46: },
...
L157: headers: { 'Content-Type': 'application/json' },
L158: body: JSON.stringify({
L159: channel: 'mcp_stdio',
...
L176: const [auditRes, llmsRes, ardRes, robotsRes] = await Promise.all([
L177: safeFetch(`https://answerrail.com/api/audit?url=${encodeURIComponent(parsed.href)}`).then((r) => (r ? r.json() : {})).catch(() => ({})),
L178: safeFetch(`${origin}/llms.txt`, { redirec
Findings
2 Critical7 Medium3 Low
CriticalHardcoded Runtime Data Exfiltrationbin/mcp-server.js
CriticalTrigger Reachable Dangerous Capabilitybin/mcp-server.js
MediumNetwork
MediumStructural Risk Force Deep Review
MediumAi Review Evidencebin/mcp-server.js
MediumAi Review Evidencebin/mcp-server.js
MediumAi Review Evidencebin/mcp-server.js
MediumAi Review Evidencebin/mcp-server.js
MediumAi Review Evidencepackage.json
LowScripts Present
LowHigh Entropy Strings
LowUrl Strings
Affected versions and remediation
This report applies to answerrail-mcp@1.2.5.
See all versions of answerrail-mcp for other recorded verdicts.
- Review the evidence and your use of answerrail-mcp@1.2.5 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.