@aiclude/security-skill 1.1.0 → 1.2.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/SKILL.md +54 -7
- package/package.json +1 -1
package/SKILL.md
CHANGED
|
@@ -1,14 +1,55 @@
|
|
|
1
|
+
---
|
|
2
|
+
name: aiclude-security-scan
|
|
3
|
+
description: Scan MCP Servers and AI Agent Skills for security vulnerabilities. 7 parallel engines detect prompt injection, tool poisoning, malware, supply chain attacks, and more.
|
|
4
|
+
tags: [security, vulnerability, scanner, mcp, ai-agent, sast, sca, malware]
|
|
5
|
+
---
|
|
6
|
+
|
|
1
7
|
# /security-scan - AIclude Security Vulnerability Scanner
|
|
2
8
|
|
|
3
9
|
Scan MCP Servers and AI Agent Skills for security vulnerabilities. Returns existing scan results instantly if available, or registers the target and triggers a new scan automatically.
|
|
4
10
|
|
|
11
|
+
## How It Works
|
|
12
|
+
|
|
13
|
+
This skill operates in **two modes**:
|
|
14
|
+
|
|
15
|
+
1. **Name-based lookup (`--name`)**: Queries the AIclude API (`https://vs-api.aiclude.com`) to check if a scan report already exists for the given package name. If found, the report is returned immediately. If not found, the target is registered and a server-side scan is triggered. **No local code is uploaded** — only the package name and metadata are sent.
|
|
16
|
+
|
|
17
|
+
2. **Local scan (`<target-path>`)**: Reads source files from the specified local directory and runs all 7 scan engines **locally on your machine**. Results are generated locally and **no code or scan artifacts are sent to any external service**. The scan is entirely offline.
|
|
18
|
+
|
|
19
|
+
## Network Behavior
|
|
20
|
+
|
|
21
|
+
| Action | Endpoint | Data Sent | Data Received |
|
|
22
|
+
|--------|----------|-----------|---------------|
|
|
23
|
+
| Name lookup | `POST https://vs-api.aiclude.com/api/v1/scan/lookup` | Package name, type | Scan report (JSON) |
|
|
24
|
+
| Scan status polling | `GET https://vs-api.aiclude.com/api/v1/scan/status/{id}` | Job ID | Scan status |
|
|
25
|
+
| Local scan | None (offline) | Nothing | Nothing |
|
|
26
|
+
|
|
27
|
+
**No source code, file contents, or scan artifacts are ever uploaded.** Name-based lookups only transmit the package name and type.
|
|
28
|
+
|
|
29
|
+
## Required Environment Variables
|
|
30
|
+
|
|
31
|
+
| Variable | Required | Description |
|
|
32
|
+
|----------|----------|-------------|
|
|
33
|
+
| `ASVS_SIGNING_SECRET` | **Yes** (for `--name` mode) | HMAC signing secret for authenticating API requests. Must be set before using name-based lookups. Without it, name-based mode will fail with an error. |
|
|
34
|
+
| `ASVS_API_URL` | No | API server URL. Defaults to `https://vs-api.aiclude.com` |
|
|
35
|
+
|
|
36
|
+
**Local scan mode** (`/security-scan ./path`) requires **no environment variables** and works fully offline.
|
|
37
|
+
|
|
38
|
+
## Authentication & Privacy
|
|
39
|
+
|
|
40
|
+
- API requests are authenticated via HMAC-SHA256 signatures using `ASVS_SIGNING_SECRET`. No user tokens or credentials are stored.
|
|
41
|
+
- Name-based lookups transmit only: package name, target type, and a timestamp-based signature.
|
|
42
|
+
- Local scans are fully offline — no network requests are made.
|
|
43
|
+
- Scan reports on the web dashboard (https://vs.aiclude.com) are public and contain only vulnerability metadata, not source code.
|
|
44
|
+
- No telemetry, analytics, or tracking data is collected.
|
|
45
|
+
|
|
5
46
|
## Usage
|
|
6
47
|
|
|
7
48
|
```
|
|
8
|
-
# Search by name (
|
|
49
|
+
# Search by name (requires ASVS_SIGNING_SECRET env var, queries remote API)
|
|
9
50
|
/security-scan --name <package-name> [options]
|
|
10
51
|
|
|
11
|
-
# Scan a local directory directly
|
|
52
|
+
# Scan a local directory directly (fully offline, no env vars needed)
|
|
12
53
|
/security-scan <target-path> [options]
|
|
13
54
|
```
|
|
14
55
|
|
|
@@ -24,13 +65,13 @@ Scan MCP Servers and AI Agent Skills for security vulnerabilities. Returns exist
|
|
|
24
65
|
## Examples
|
|
25
66
|
|
|
26
67
|
```
|
|
27
|
-
# Search security scan results for an MCP server
|
|
68
|
+
# Search security scan results for an MCP server (remote lookup)
|
|
28
69
|
/security-scan --name @anthropic/mcp-server-fetch
|
|
29
70
|
|
|
30
|
-
# Search scan results for a Skill
|
|
71
|
+
# Search scan results for a Skill (remote lookup)
|
|
31
72
|
/security-scan --name my-awesome-skill --type skill
|
|
32
73
|
|
|
33
|
-
# Scan local source code
|
|
74
|
+
# Scan local source code (fully offline)
|
|
34
75
|
/security-scan ./my-mcp-server
|
|
35
76
|
```
|
|
36
77
|
|
|
@@ -65,6 +106,12 @@ The report includes:
|
|
|
65
106
|
3. **Risk Assessment** (what risks are present and their impact)
|
|
66
107
|
4. **Remediation Recommendations** (how to fix each vulnerability)
|
|
67
108
|
|
|
68
|
-
##
|
|
109
|
+
## Source Code & npm Package
|
|
110
|
+
|
|
111
|
+
- **npm**: [`@aiclude/security-skill`](https://www.npmjs.com/package/@aiclude/security-skill)
|
|
112
|
+
- **MCP Server**: [`@aiclude/security-mcp`](https://www.npmjs.com/package/@aiclude/security-mcp)
|
|
113
|
+
- **Web Dashboard**: https://vs.aiclude.com
|
|
114
|
+
|
|
115
|
+
## License
|
|
69
116
|
|
|
70
|
-
|
|
117
|
+
MIT - AICLUDE Inc.
|
package/package.json
CHANGED