App framework, testing framework, and inspector for MCP Apps.
- capability exposure inferred + 22
- tool safety inferred + 12
- trust mitigators mixed − 14
inferred mixed
The A–E grade is our heuristic synthesis — a "review this" prompt, not a verdict. Each factor is tagged by what backs it: attested (a verifiable record), reported (a third party's claim), or inferred (our own heuristic, e.g. permissions). See methodology.
grade last moved 1w ago · see ecosystem CVEs →
- C · 35 → B · 20
- C · 55 → C · 35
- C · 48 → C · 55
- C · 39 → C · 48
- C · 44 → C · 39
- C · 51 → C · 44
- C · 56 → C · 51
- A · 0 → C · 56
No known CVEs for this server.
- high dangerous code
dynamic exec: new Function() · obfuscated payload: dynamic require()/import()
analyzed v0.20.74 · analyzer v33 · 4w ago
skills & prompt files 3
danger signals6
- dynamic code execution new Function() Alignbase-sunpeak-44f8ca2/packages/sunpeak/bin/lib/extract-resource.mjs :52
new Function('module', 'exports', 'require', code)(mod, mod.exports, () => ({})); - dynamic code execution new Function() Alignbase-sunpeak-44f8ca2/packages/sunpeak/bin/lib/extract-tool.mjs :59
const fn = new Function(code + '\nreturn tool;'); - dynamic code execution new Function() Alignbase-sunpeak-44f8ca2/packages/sunpeak/src/inspector/inline-helper-runtime.test.ts :41
const fn = new Function('window', 'document', SUNPEAK_INLINE_HELPER_SCRIPT); - dynamic code execution new Function() Alignbase-sunpeak-44f8ca2/packages/sunpeak/src/lib/extract-resource.ts :54
new Function('module', 'exports', 'require', code)(mod, mod.exports, () => ({})); - dynamic code execution new Function() Alignbase-sunpeak-44f8ca2/packages/sunpeak/src/lib/extract-tool.ts :64
const fn = new Function(code + '\nreturn tool;'); - suspicious endpoint 169.254.169.254 (cloud metadata) Alignbase-sunpeak-44f8ca2/packages/sunpeak/src/cli/inspect.test.ts :442
_securityTestExports.assertHttpServerUrlAllowed('http://169.254.169.254/latest')
[](https://mcpobservatory.com/servers/github:Sunpeak-AI/sunpeak/security) Heuristic, inferred signals — false positives (legitimately powerful tools, forks, language ports) are expected. Treat each as "review this", not a verdict. See the ecosystem-wide picture on the security hub, or the fleet security of Sunpeak-AI.