First Seen
Feb 18, 2026
Last Scanned
Feb 22, 2026
Findings
4
Score
100/100
Findings (4)
Detects URLs fetched at runtime that control or influence agent behavior without pinning
GET https:// + SettingsAvoid loading configuration or behavior-controlling content from runtime URLs. Bundle required configurations locally or pin remote config to versioned, integrity-verified endpoints.
Likely FP if the URL in the match is a documentation link or example URL (e.g., example.com) rather than an actual runtime-fetched configuration endpoint.
Detects MCP server configurations connecting to non-localhost remote URLs
"url": "https://example.com/article"Change the MCP server URL to localhost or a trusted internal endpoint. If a remote server is required, verify the domain ownership and use HTTPS with certificate validation.
Likely FP if the URL points to example.com, a documentation domain, or a well-known SaaS API endpoint (e.g., api.openai.com).
Detects MCP server configurations connecting to non-localhost remote URLs
"url": "https://example.com"Change the MCP server URL to localhost or a trusted internal endpoint. If a remote server is required, verify the domain ownership and use HTTPS with certificate validation.
Likely FP if the URL points to example.com, a documentation domain, or a well-known SaaS API endpoint (e.g., api.openai.com).
Detects eval() or exec() used for dynamic code execution
exec('Replace eval()/exec() with a safer alternative such as json.loads(), ast.literal_eval(), or a purpose-built parser.
Likely FP if the matched text contains 'exec' as part of a word (e.g., 'execute', 'execution') rather than an actual eval() or exec() call.