sandspy 0.1.1

Real-time security monitor for AI coding agents
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

[agent]
name = "Cursor"
process_names = ["cursor", "cursor-helper"]
description = "Anysphere's AI-powered code editor"

[expected.network]
allowed_domains = [
  "api2.cursor.sh", "api3.cursor.sh", "api4.cursor.sh", "api5.cursor.sh",
  "repo42.cursor.sh", "cursor-cdn.com", "marketplace.cursorapi.com",
  "authenticate.cursor.sh", "*.gcpp.cursor.sh",
  "telemetry.cursor.sh", "*.statsig.com",
]

[expected.filesystem]
normal_patterns = [
  "**/*.rs", "**/*.py", "**/*.ts", "**/*.js", "**/*.tsx", "**/*.jsx",
  "**/Cargo.toml", "**/package.json", "**/tsconfig.json",
  "**/.git/**", "**/README*", "**/LICENSE*",
]

[alerts]
sensitive_file_access = true
unknown_network = true
shell_dangerous_commands = true
clipboard_read = true
env_secret_access = true
excessive_file_reads = 500
excessive_data_out = "10MB"

[risk_weights]
sensitive_file_read = 15
unknown_network_connection = 25
secret_env_access = 20
dangerous_command = 30
clipboard_read = 10
excessive_scope = 15