registry / Credential exfiltration / @testsmith/api-spector-agent
@testsmith/api-spector-agent
API Spector private runner: monitor internal/local APIs the cloud cannot reach, with no inbound connections.
This report shows the evidence for @testsmith/api-spector-agent@0.9.6. 1 of 35 known versions is blocked or flagged. The newest version, 0.9.9, is allowed. See all versions.
AI Security Review
scanned 22h ago · by lpm-firewall-aiThe runner forwards authentication headers from monitored APIs in its routine result uploads to APP_URL. Although result posting is part of the advertised service, the source does not show redaction of these third-party credentials before submission.
Static reason
One or more suspicious static signals were detected.
Trigger
The agent starts and polls monitors automatically after launch.
Impact
Credentials for monitored APIs may be disclosed to the API Spector service.
Mechanism
It adds monitor credentials to tested requests, copies the sent request headers into sent_headers, then posts those headers to APP_URL.
Attack narrative
When a monitor uses bearer or API-key authentication, the runner places that credential in the request headers sent to the monitored API. It then copies the sent headers into sent_headers and posts that field to APP_URL as part of the result. This happens during the runner's automatic polling cycle, so credentials for monitored services can be sent to the API Spector service without a separate per-monitor disclosure or confirmation.
AI rationale
The runner forwards authentication headers from monitored APIs in its routine result uploads to APP_URL. Although result posting is part of the advertised service, the source does not show redaction of these third-party credentials before submission.
Evidence
dist/index.cjs
Network endpoints2
${APP_URL}/api/monitor-runtime/due${APP_URL}/api/monitor-runtime/{m6.id}/result
Decision evidence
public snapshotAI called this Malicious at 94.0% confidence as Malware with low false-positive risk.
Evidence for block
- The manifest exposes dist/index.cjs as both the CLI and package main; the entry point starts polling when run.
- The bundled monitor engine puts monitor bearer tokens and API keys into outgoing request headers.
- The agent copies sent request headers into sent_headers and includes them in results posted to APP_URL.
- Polling and result submission run automatically at startup and repeat on a timer.
Evidence against
- The package is described as a private API monitor that runs checks inside the user's network.
- The README says results are posted to API Spector Cloud, and the user supplies APP_URL and an agent token.
Behavioral surface
CryptoEnvironmentVarsFilesystemNetwork
HighEntropyStringsPossibleObfuscationUrlStrings
Oversized source lightweight scan
dist/index.cjs6.13 MB file, sampled 256 KB
FilesystemNetworkEnvironmentVarsCryptoPossibleObfuscationHighEntropyStringsUrlStringsdeveloper.mozilla.orgfeross.orggithub.comjimmy.warting.senodejs.orgsocket.iowww.w3.org
Source & flagged code
8 flaggeddist/index.cjsView file
•Manifest-reachable source captures an API credential, sends it to a fixed unofficial gateway, and persists that redirection in agent or shell configuration.
dist/index.cjs:
"Authorization",
"Proxy-Authorization",
get baseUrl() {
return this.baseUrl?.origin;
this[kProxyHeaders]["proxy-authorization"] = `Basic ${opts.auth}`;
this[kProxyHeaders]["proxy-authorization"] = opts.token;
this[kProxyHeaders]["proxy-authorization"] = `Basic ${Buffer.from(`${decodeURIComponent(username)}:${decodeURIComponent(password)}`).toString("base64")}`;
const existProxyAuth = headers && Object.keys(headers).find((key) => key.toLowerCase() === "proxy-authorization");
High
Credential Redirect Persistence
Manifest-reachable source sends a prompted API credential to a fixed unofficial gateway and persists the redirection.
dist/index.cjsView on unpkg247}, m7.utcOffset = function() {
L248: return 15 * -Math.round(this.$d.getTimezoneOffset() / 15);
L249: }, m7.diff = function(r64, d9, l9) {
...
L356: * @see DOMParser.SupportedType.isHTML
L357: * @see https://www.iana.org/assignments/media-types/text/html IANA MimeType registration
L358: * @see https://en.wikipedia.org/wiki/HTML Wikipedia
...
L586: /**
L587: * @private
L588: * @param {function (Node):boolean} predicate
...
L993: function _xmlEncoder(c8) {
L994: return c8 == "<" && "<" || c8 == ">" && ">" || c8 == "&" && "&" || c8 == '"' && """ || "&#" + c8.charCodeAt() + ";";
L995: }
High
Command Output Exfiltration
Source combines command execution, command-output handling, and outbound requests; review data flow before blocking.
dist/index.cjsView on unpkg · L247247}, m7.utcOffset = function() {
L248: return 15 * -Math.round(this.$d.getTimezoneOffset() / 15);
L249: }, m7.diff = function(r64, d9, l9) {
...
L356: * @see DOMParser.SupportedType.isHTML
L357: * @see https://www.iana.org/assignments/media-types/text/html IANA MimeType registration
L358: * @see https://en.wikipedia.org/wiki/HTML Wikipedia
...
L586: /**
L587: * @private
L588: * @param {function (Node):boolean} predicate
...
L993: function _xmlEncoder(c8) {
L994: return c8 == "<" && "<" || c8 == ">" && ">" || c8 == "&" && "&" || c8 == '"' && """ || "&#" + c8.charCodeAt() + ";";
L995: }
High
Obfuscated Payload Loader
Source contains an obfuscated payload loader that reconstructs and executes hidden code.
dist/index.cjsView on unpkg · L247•stage = ast_semantic_analysis; reason = ast_trace_path_limit_exceeded; limitedFiles = 1
High
Semantic Analysis Limited
A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
dist/index.cjsView on unpkg247}, m7.utcOffset = function() {
L248: return 15 * -Math.round(this.$d.getTimezoneOffset() / 15);
L249: }, m7.diff = function(r64, d9, l9) {
...
L356: * @see DOMParser.SupportedType.isHTML
L357: * @see https://www.iana.org/assignments/media-types/text/html IANA MimeType registration
L358: * @see https://en.wikipedia.org/wiki/HTML Wikipedia
...
L586: /**
L587: * @private
L588: * @param {function (Node):boolean} predicate
...
L993: function _xmlEncoder(c8) {
L994: return c8 == "<" && "<" || c8 == ">" && ">" || c8 == "&" && "&" || c8 == '"' && """ || "&#" + c8.charCodeAt() + ";";
L995: }
Medium
Unsafe Vm Context
Package source executes code through a VM context API.
dist/index.cjsView on unpkg · L247•path = dist/index.cjs
kind = oversized_source_file
sizeBytes = 6430451
magicHex = [redacted]
Medium
Oversized Source File
Package contains source files above the normal full-analysis size ceiling.
dist/index.cjsView on unpkg•path = dist/index.cjs
kind = oversized_cli_entrypoint
sizeBytes = 6430451
magicHex = [redacted]
Medium
Oversized Cli Entrypoint
Package contains an oversized executable-looking CLI entrypoint.
dist/index.cjsView on unpkg247}, m7.utcOffset = function() {
L248: return 15 * -Math.round(this.$d.getTimezoneOffset() / 15);
L249: }, m7.diff = function(r64, d9, l9) {
...
L356: * @see DOMParser.SupportedType.isHTML
L357: * @see https://www.iana.org/assignments/media-types/text/html IANA MimeType registration
L358: * @see https://en.wikipedia.org/wiki/HTML Wikipedia
...
L586: /**
L587: * @private
L588: * @param {function (Node):boolean} predicate
...
L993: function _xmlEncoder(c8) {
L994: return c8 == "<" && "<" || c8 == ">" && ">" || c8 == "&" && "&" || c8 == '"' && """ || "&#" + c8.charCodeAt() + ";";
L995: }
Low
Weak Crypto
Package source references weak cryptographic algorithms.
dist/index.cjsView on unpkg · L247Findings
4 High6 Medium4 Low
HighCredential Redirect Persistencedist/index.cjs
HighCommand Output Exfiltrationdist/index.cjs
HighObfuscated Payload Loaderdist/index.cjs
HighSemantic Analysis Limiteddist/index.cjs
MediumUnsafe Vm Contextdist/index.cjs
MediumNetwork
MediumEnvironment Vars
MediumOversized Source Filedist/index.cjs
MediumOversized Cli Entrypointdist/index.cjs
MediumStructural Risk Force Deep Review
LowWeak Cryptodist/index.cjs
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
Affected versions and remediation
This report applies to @testsmith/api-spector-agent@0.9.6.
The versions list below shows the verdict for each recorded version.
- Avoid installing @testsmith/api-spector-agent@0.9.6. 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.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.
Versions
35 known versions0.9.9latestClean—AllowPublishedNov 4, 2026, 1:58 PM0.9.7Clean—AllowPublishedNov 4, 2026, 9:49 AM0.9.6shownMaliciousCredential ExfiltrationBlockPublishedNov 3, 2026, 10:21 PM0.9.3Clean—AllowPublishedNov 3, 2026, 8:41 AM0.9.1Clean—AllowPublishedOct 29, 2026, 8:01 PM0.9.0Clean—AllowPublishedOct 28, 2026, 10:07 PM0.8.9Clean—AllowPublishedOct 28, 2026, 7:50 AM0.8.7Clean—AllowPublishedOct 26, 2026, 4:03 PM0.8.6Clean—AllowPublishedOct 26, 2026, 12:14 PM0.8.5Clean—AllowPublishedOct 26, 2026, 11:32 AM0.8.4Clean—AllowPublishedOct 23, 2026, 6:31 AM0.8.1Clean—AllowPublishedOct 20, 2026, 7:51 AM0.8.0Clean—AllowPublishedOct 19, 2026, 1:55 PM0.7.9Clean—AllowPublishedOct 19, 2026, 12:44 PM0.7.8Clean—AllowPublishedOct 19, 2026, 10:36 AM0.7.7Clean—AllowPublishedOct 14, 2026, 6:51 PM0.7.6SuspiciousRemote Code ExecutionWarnPublishedOct 13, 2026, 9:18 AM0.7.5Clean—AllowPublishedOct 12, 2026, 9:23 AM0.7.4Clean—AllowPublishedOct 7, 2026, 4:39 PM0.7.3Clean—AllowPublishedOct 6, 2026, 8:21 PM0.7.2SuspiciousRemote Code ExecutionWarnPublishedSep 30, 2026, 2:49 PM0.7.1Clean—AllowPublishedSep 28, 2026, 8:49 PM0.7.0Clean—AllowPublishedSep 28, 2026, 4:26 PM0.6.4SuspiciousCredential ExfiltrationWarnPublishedSep 28, 2026, 9:38 AM0.6.3Clean—AllowPublishedSep 27, 2026, 1:26 PM0.6.2Clean—AllowPublishedSep 27, 2026, 8:33 AM0.6.1Clean—AllowPublishedSep 25, 2026, 12:10 PM0.5.8Clean—AllowPublishedSep 23, 2026, 6:45 PM0.5.4Clean—AllowPublishedSep 22, 2026, 5:32 PM0.5.3Clean—AllowPublishedSep 13, 2026, 11:16 AM0.5.2SuspiciousData ExfiltrationWarnPublishedSep 8, 2026, 7:59 PM0.5.1SuspiciousData ExfiltrationWarnPublishedSep 7, 2026, 8:46 AM0.5.0SuspiciousRemote Code ExecutionWarnPublishedSep 5, 2026, 4:12 PM0.4.9Clean—AllowPublishedAug 27, 2026, 8:44 PM0.4.8SuspiciousCredential ExfiltrationWarnPublishedAug 24, 2026, 9:33 PM