superglue (YC W25) builds integrations and tools from natural language. Get production-grade tools for long tail and enterprise systems.
- capability exposureinferred+35
- recent driftinferred+12
- tool safetyinferred+12
inferred
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.
graded 10m ago · see ecosystem CVEs →
- A · 0 → C · 59
No known CVEs for this server.
- highdangerous code
dynamic exec: eval(), new Function(), vm exec
analyzed commit fc05430 · analyzer v18 · 6h ago
skills & prompt files 3
danger signals12
- dynamic code executioneval()superglue-ai-superglue-fc05430/eval/llm-benchmark-evals/services/code-executor.ts:21
const evalResult = eval(wrappedCode); - dynamic code executionnew Function()superglue-ai-superglue-fc05430/packages/core/auth/access-rule-evaluator.ts:142
const fn = new Function("stepConfig", `return Boolean(${expression})`); - dynamic code executioneval()superglue-ai-superglue-fc05430/packages/core/deno-runtime/strategies/http.ts:129
const fn = eval(code); - dynamic code executioneval()superglue-ai-superglue-fc05430/packages/core/deno-runtime/utils/transform.ts:75
fn = eval(wrappedCode); - dynamic code executionnew Function()superglue-ai-superglue-fc05430/packages/shared/utils/vm-helpers.ts:14
const fn = new Function("sourceData", wrappedCode); - dynamic code executionvm execsuperglue-ai-superglue-fc05430/packages/web/src/lib/agent/registries/tool-registry.ts:3032
const allowed = vm.runInNewContext(`"use strict"; Boolean(${input.expression})`, sandbox, { - suspicious endpointapi.telegram.orgsuperglue-ai-superglue-fc05430/packages/shared/templates.ts:3018
apiUrl: "https://api.telegram.org", - over-broad OAuth scopeadmin:orgsuperglue-ai-superglue-fc05430/packages/shared/templates.ts:307
"repo user admin:org workflow gist notifications delete_repo write:packages read:packages", - over-broad OAuth scopehttps://mail.google.com/superglue-ai-superglue-fc05430/packages/shared/templates.ts:467
scopes: "https://mail.google.com/", - over-broad OAuth scopehttps://www.googleapis.com/auth/drivesuperglue-ai-superglue-fc05430/packages/shared/templates.ts:505
scopes: "https://www.googleapis.com/auth/drive", - over-broad OAuth scopehttps://www.googleapis.com/auth/spreadsheetssuperglue-ai-superglue-fc05430/packages/shared/templates.ts:566
scopes: "https://www.googleapis.com/auth/spreadsheets https://www.googleapis.com/auth/drive", - over-broad OAuth scopehttps://www.googleapis.com/auth/cloud-platformsuperglue-ai-superglue-fc05430/packages/shared/templates.ts:681
scopes: "https://www.googleapis.com/auth/cloud-platform",
- recent drift+12 capability drift →
[](https://mcpobservatory.com/servers/github:superglue-ai/superglue/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 superglue-ai.