Code analysis

static source read inferred

Static code-analysis findings — hidden prompt content in shipped skill files, committed secrets, dynamic-exec sinks, and suspicious call-home endpoints — across the analyzed catalogue. Heuristic, pure, no code executed; every row deep-links to its source. Click a kind to filter.

analysis coverage 86% of 59,713 analyzable servers
51314 analyzed
7403 re-analysis due
996 not analyzable
0 not yet analyzed
4850 source gone
not analyzable
796 too large 200 no source

Running analyzer v33. The scanner changelog explains what each version detects and when it changed.

code findings 15 shown
  1. HIGH dynamic exec ashlrai/webfetch new Function() let templateFn = new Function(`return ${template}`)();
  2. HIGH dynamic exec andrewLi1994/codex-gemini-web-bridge new Function() const makeValidate = new Function(`${names_1.default.self}`, `${names_1.default.scope}`, sourceCode);
  3. MEDIUM suspicious endpoint drunkrhin0/rapid7-docs-mcp 169.254.169.254 (cloud metadata) expect(isValidProductUrl('https://169.254.169.254/latest/meta-data/')).toBe(false);
  4. MEDIUM over-broad oauth scope crmy-ai/crmy https://mail.google.com/ ? has('https://www.googleapis.com/auth/gmail.send') || has('https://mail.google.com/')
  5. MEDIUM over-broad oauth scope crmy-ai/crmy https://mail.google.com/ 'https://mail.google.com/',
  6. MEDIUM suspicious endpoint off-grid-ai/desktop developer.mixpanel.com (telemetry) docsUrl: 'https://developer.mixpanel.com/',
  7. MEDIUM suspicious endpoint off-grid-ai/desktop amplitude.com (telemetry) docsUrl: 'https://amplitude.com/docs',
  8. MEDIUM suspicious endpoint off-grid-ai/desktop posthog.com (telemetry) docsUrl: 'https://posthog.com/docs',
  9. MEDIUM suspicious endpoint off-grid-ai/desktop mcp.posthog.com (telemetry) url: 'https://mcp.posthog.com/sse',
  10. MEDIUM suspicious endpoint off-grid-ai/desktop docs.sentry.io (telemetry) docsUrl: 'https://docs.sentry.io/product/sentry-mcp/',
  11. HIGH obfuscation off-grid-ai/desktop dynamic require()/import() const runtime = await import(`${pathToFileURL(runtimeEntry).href}?contract-check=${Date.now()}`)
  12. MEDIUM suspicious endpoint cognis-digital/toolguard 169.254.169.254 (cloud metadata) d = self.guard.check("http_get", {"url": "http://169.254.169.254/meta"})
  13. HIGH committed secret aldok10/zara-jira-mcp Slack token xoxb-y…(19 chars, redacted)
  14. MEDIUM suspicious endpoint LinChuang2008/nightmend api.telegram.org api_url = f"https://api.telegram.org/bot{bot_token}/sendMessage"
  15. HIGH credential in log unison-labs-ai/claude-unison credential in log console.log(`Token (unverified): ${result.apiKey}`);