Malinois

Check a deployed app for data leaks

scan_app

Idempotent

Runs a passive, outside-in security check of a live web app and returns a letter grade (A–F), each issue in plain language with fix steps, and a report link. Use when the user asks whether their deployed app is safe, before launch, or after a redeploy to confirm a fix. It checks for publicly readable Supabase/Firebase data, secret keys (Stripe, OpenAI, Supabase service_role…) in client JavaScript, downloadable .env/.git files, source maps, permissive CORS and missing security headers. Do not use it for apps the user does not own or is not authorized to test, for localhost or private addresses, or to review source code — it only sees what the public URL serves. Behavior: sends ordinary GET requests like a browser (no login, exploitation or load testing); takes about 10–30 seconds; saves the result as a report page on malinois.app, linked in the response; secrets appear only masked. Each app can be checked at most 20 times per hour.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "url": "<url>",
      "i_own_this": "<i_own_this>"
    },
    "name": "scan_app"
  }
}

Input Schema

Name	Required	Description
`url`	Yes	Public http(s) address of the deployed app, e.g. https://my-app.lovable.app (scheme optional).
`lang`	No	Language for the explanations (default: en).
`i_own_this`	Yes	Must be true. Set it only after the user has explicitly confirmed they own this app or are authorized to test it; without it the check is refused.

Output Schema

Name	Required	Description
`host`	Yes
`grade`	Yes	A (best) to F
`score`	Yes	0–100
`limited`	Yes	True when the app exposed little to a passive check; a good grade is then not proof of safety.
`findings`	Yes	Most serious first.
`platform`	No	Detected builder/host, e.g. lovable, replit
`report_url`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

Changed4 schema fields changedSep 11, 2026

changedInput schema / properties / i_own_this / description

Previous value: -"Set to true only after the user has confirmed they own this app or are authorized to test it."New value: +"Must be true. Set it only after the user has explicitly confirmed they own this app or are authorized to test it; without it the check is refused."

changedInput schema / properties / url / description

Previous value: -"Public URL of the deployed app, e.g. https://my-app.lovable.app"New value: +"Public http(s) address of the deployed app, e.g. https://my-app.lovable.app (scheme optional)."

addedInput schema / properties / url / maxLength
```
Added value: +500
```

changedOutput schema / (root)

Previous value: -nullNew value: +{
+  "properties": {
+    "findings": {
+      "description": "Most serious first.",
+      "items": {
+        "properties": {
+          "evidence": {
+            "description": "Masked evidence; secrets are never returned in full.",
+            "type": [
+              "string",
+              "null"
+            ]
+          },
+          "rule_id": {
+            "type": "string"
+          },
+          "severity": {
+            "enum": [
+              "critical",
+              "high",
+              "medium",
+              "low",
+              "info"
+            ],
+            "type": "string"
+          },
+          "title": {
+            "type": "string"
+          },
+          "what_it_means": {
+            "type": [
+              "string",
+              "null"
+            ]
+          },
+          "what_to_do": {
+            "type": [
+              "string",
+              "null"
+            ]
+          }
+        },
+        "required": [
+          "rule_id",
+          "severity",
+          "title"
+        ],
+        "type": "object"
+      },
+      "type": "array"
+    },
+    "grade": {
+      "description": "A (best) to F",
+      "type": "string"
+    },
+    "host": {
+      "type": "string"
+    },
+    "limited": {
+      "description": "True when the app exposed little to a passive check; a good grade is then not proof of safety.",
+      "type": "boolean"
+    },
+    "platform": {
+      "description": "Detected builder/host, e.g. lovable, replit",
+      "type": [
+        "string",
+        "null"
+      ]
+    },
+    "report_url": {
+      "type": "string"
+    },
+    "score": {
+      "description": "0–100",
+      "type": "number"
+    }
+  },
+  "required": [
+    "grade",
+    "score",
+    "host",
+    "limited",
+    "report_url",
+    "findings"
+  ],
+  "type": "object"
+}

First observedSep 11, 2026

TDQS

A4.7/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only declare safety flags; the description adds substantial operational context beyond them: ordinary GET requests with no login/exploitation/load testing, a 10–30 second runtime, a persisted report page on malinois.app linked in the response, secrets shown only masked, and a 20-per-hour rate limit. The readOnlyHint=false annotation is consistent with the disclosed side effect of writing a report page.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with the core purpose in the first sentence, then usage, then exclusions, then behavior. Three tight paragraphs, no redundancy, and every sentence carries distinct information an agent needs.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Although an output schema exists, the description still summarizes the return shape (grade, issues with fixes, report link). Combined with the consent prerequisite, authorization limits, runtime, rate limit, and scope boundaries, an agent has everything needed to decide and invoke correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already documents url, lang, and i_own_this, including the consent semantics and refusal behavior. The description reinforces the ownership/authorization constraint but adds no parameter-level detail (formats, defaults, enum meaning) beyond the schema, so the baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description gives a specific verb and resource ('runs a passive, outside-in security check of a live web app') plus the exact outputs (letter grade A–F, plain-language issues with fix steps, report link). It also enumerates what is checked (public Supabase/Firebase data, leaked secrets, .env/.git, source maps, CORS, headers), making it unmistakably distinct from the sibling explain_finding.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicit triggers are given ('when the user asks whether their deployed app is safe, before launch, or after a redeploy to confirm a fix') and explicit exclusions are listed (apps the user doesn't own or isn't authorized to test, localhost/private addresses, source-code review). This is exactly the when/when-not guidance the dimension asks for.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All Malinois tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?