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.
50020 analyzed
8131 re-analysis due
988 not analyzable
1 not yet analyzed
4803 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.
- hidden prompt 338
- committed secret 6177
- dynamic exec 11826
- obfuscation 3575
- suspicious endpoint 12346
- credential in log 713
- over-broad oauth scope 2548
- suspicious skill script 185
- bundled IDE extension 49
- skill file 187620
- MEDIUM suspicious endpoint fouradata/mcp 8.8.8.8
test("3. 8.8.8.8 (Google DNS)", async () => assert.equal(await isBlocked("https://8.8.8.8"), false)); - MEDIUM suspicious endpoint fouradata/mcp 1.1.1.1
test("2. 1.1.1.1 (Cloudflare)", async () => assert.equal(await isBlocked("https://1.1.1.1"), false)); - MEDIUM suspicious endpoint fouradata/mcp 1.1.1.1
const target = { method: "GET", url: "https://1.1.1.1/" }; - MEDIUM suspicious endpoint fouradata/mcp 1.1.1.1
const target = { method: "GET", url: "https://1.1.1.1/" }; - MEDIUM suspicious endpoint fouradata/mcp 169.254.169.254 (cloud metadata)
aws_metadata: "http://169.254.169.254/latest/meta-data/", - MEDIUM suspicious endpoint fouradata/mcp 172.32.0.0
just_outside_172: "http://172.32.0.0", // public, must NOT block - HIGH dynamic exec awarselabs/awarse-mcp eval()/exec()
locator = eval(py_expression, {"page": page}) - HIGH dynamic exec awarselabs/awarse-mcp eval()/exec()
healed_locator = eval(resolved_call, {"page": self._page}) - HIGH dynamic exec editmamei/editmamei new Function()
return new Function(`${source}\nreturn __mcpJsonEncode;`)() as (v: unknown) => string; - HIGH dynamic exec editmamei/editmamei new Function()
const fn = new Function( - MEDIUM suspicious endpoint beepboop2025/liquilens-site t.me
assert.equal(destination.origin + destination.pathname, "https://t.me/liquilens_crypto_bot"); - MEDIUM suspicious endpoint beepboop2025/liquilens-site t.me
assert.equal(bot.origin, "https://t.me"); - HIGH dynamic exec beepboop2025/liquilens-site vm exec
vm.runInNewContext(script, context); - MEDIUM suspicious endpoint beepboop2025/liquilens-site t.me
return `https://t.me/${bot}?start=social_${source}_${task}`; - MEDIUM suspicious endpoint beepboop2025/liquilens-site t.me
var destination = "https://t.me/liquilens_crypto_bot?start=" + encodeURIComponent(start);