Analyze command
The analyze command is used to analyze a file the output files of JS recon modules to determine if they contain sensitive info or not.
Usage
js-recon analyze [options]
Options
| Option | Alias | Description | Default | Required |
|---|---|---|---|---|
--rules <file/dir> | -r | Rules file or directory | No | |
--mapped-json <file> | -m | Mapped JSON file | mapped.json | No |
--tech <tech> | -t | Technology used in the JS files (run with -l/--list to see available options) | No | |
--openapi <file> | Path to OpenAPI spec file | No | ||
--list | -l | List available technologies | false | No |
--validate | Validate the rules | false | No | |
--output <file> | -o | Output JSON file name | analyze.json | No |
--disable-rules-version-check | Skip the GitHub rules version check and use cached rules as-is | false | No |
Examples
Validating custom rules
Though the tool will validate all the templates automatically, you can also validate them manually using the --validate option.
Assuming that the rules are stored in the ./rules directory, you can run the following command to validate them:
js-recon analyze --validate -r ./rules
Analyzing OpenAPI spec file
To run analysis on an OpenAPI spec file, make sure that you've generated it using the map command:
js-recon map -d output/<domain> -t <tech> --openapi
This will give a file called mapped-openapi.json in the working directory. To run analysis on this, run the following command:
js-recon analyze --openapi mapped-openapi.json -t <tech>
This will analyze the mapped-openapi.json file against the default rules that would be stored in $HOME/.js-recon/rules.
Built-in detection rules
The bundled rules detect both request-level misconfigurations (for example, missing Authorization header) and client-side code patterns such as DOM XSS sinks and Client-Side Path Traversal (CSPT). See Predefined rules for the full catalogue and the constraints each rule applies to avoid false positives.
Output format
analyze.json is a JSON array where each element represents one finding. An empty array means no rules fired.
[
{
"ruleId": "dom_xss_innerHTML",
"ruleName": "DOM XSS via innerHTML",
"ruleType": "ast",
"ruleDescription": "Detects direct assignment to innerHTML from a user-controlled source.",
"ruleAuthor": "js-recon",
"ruleTech": ["next", "react"],
"severity": "high",
"message": "[+] \"DOM XSS via innerHTML\" found in chunk 4821",
"findingLocation": "// 4821\n\ndocument.getElementById('out').innerHTML = userInput"
}
]
Fields
| Field | Type | Description |
|---|---|---|
ruleId | string | Unique rule identifier from the YAML rule file. |
ruleName | string | Human-readable rule name. |
ruleType | string | Engine that produced the finding: ast, request, or cs-mast-s. |
ruleDescription | string | Description of what the rule detects. |
ruleAuthor | string | Author of the rule. |
ruleTech | string[] | Technologies the rule targets. Values: next, vue, react, svelte, angular, or all. |
severity | string | Finding severity: info, low, medium, or high. |
message | string | One-line summary of the finding (see below). |
findingLocation | string | Where the finding was detected (see below). |
message format
The message field is a one-line human-readable label:
- AST and CS-MAST-S rules:
[+] "<rule name>" found in chunk <chunk id> - Request rules:
[+] "<rule name>" found in <path> [<METHOD>]
findingLocation format
The findingLocation field varies by engine:
-
AST rules: A comment with the chunk ID followed by the matched source code snippet:
// <chunk id>
<matched source code> -
CS-MAST-S rules: The chunk ID and matched CS-MAST-S signature(s):
// chunk: <chunk id>
// CS-MAST-S signature: <phc string> -
Request rules: The matched endpoint path and HTTP method:
<path> [<METHOD>]