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 85% of 59,163 analyzable servers
50108 analyzed
8067 re-analysis due
988 not analyzable
0 not yet analyzed
4804 source gone
not analyzable
789 too large 199 no source

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

code findings 15 shown
  1. MEDIUM suspicious endpoint alvinjchoi/opensource-loyalty 169.254.169.254 (cloud metadata) url: "https://169.254.169.254/latest/meta-data",
  2. MEDIUM suspicious endpoint alvinjchoi/opensource-loyalty 169.254.169.254 (cloud metadata) "https://169.254.169.254/latest/meta-data",
  3. HIGH suspicious skill script SunMe1977/HermesClawZero-ConfigSidecar suspicious bundled script python -c "import requests" 2>$null
  4. HIGH suspicious skill script SunMe1977/HermesClawZero-ConfigSidecar suspicious bundled script python -c "import requests" 2>$null
  5. HIGH dynamic exec SunMe1977/HermesClawZero-ConfigSidecar __import__ sink maxsize=int(__import__("os").getenv("RATE_LIMIT_MAX_IPS", "10000")), ttl=60
  6. HIGH dynamic exec SunMe1977/HermesClawZero-ConfigSidecar __import__ sink __import__("os").getenv("LOG_LEVEL", "INFO").upper(), logging.INFO
  7. MEDIUM suspicious endpoint SunMe1977/HermesClawZero-ConfigSidecar api.telegram.org url = f"https://api.telegram.org/bot{TELEGRAM_BOT_TOKEN}/sendMessage"
  8. HIGH committed secret zent7x/mcp-guard GitHub token ghp_01…(40 chars, redacted)
  9. MEDIUM suspicious endpoint enconvert/mcp us.i.posthog.com (telemetry) export const POSTHOG_HOST = process.env.POSTHOG_HOST?.trim() || "https://us.i.posthog.com";
  10. MEDIUM suspicious endpoint frangelbarrera/ARNES 169.254.169.254 (cloud metadata) result = await tool.execute({"url": "http://169.254.169.254/latest/meta-data/"}, ctx)
  11. MEDIUM suspicious endpoint frangelbarrera/ARNES 8.8.8.8 err, ip = await _check_ssrf_async("http://8.8.8.8/")
  12. MEDIUM suspicious endpoint frangelbarrera/ARNES 169.254.169.254 (cloud metadata) err, ip = await _check_ssrf_async("http://169.254.169.254/latest/meta-data/")
  13. HIGH dynamic exec frangelbarrera/ARNES eval()/exec() exec(completion, namespace) # noqa: S102 # nosec B102 - HumanEval runner pattern
  14. HIGH obfuscation 0x4m4/hexstrike-ai base64 decode → exec sink exec(base64.b64decode('{base64.b64encode(code.encode()).decode()}'))
  15. MEDIUM suspicious endpoint 0x4m4/hexstrike-ai 169.254.169.254 (cloud metadata) {"name": "Cloud Metadata", "payloads": ["http://169.254.169.254/latest/meta-data/"]},